Professional Documents
Culture Documents
25814-710-TSG Ran Physical Layer PDF
25814-710-TSG Ran Physical Layer PDF
0 (2006-09)
Technical Report
The present document has been developed within the 3rd Generation Partnership Project (3GPP TM) and may be further elaborated for the purposes of 3GPP.
The present document has not been subject to any approval process by the 3GPP Organizational Partners and shall not be implemented.
This Specification is provided for future development work within 3GPP only. The Organizational Partners accept no liability for any use of this Specification.
Specifications and reports for implementation of the 3GPP TM system should be obtained via the 3GPP Organizational Partners' Publications Offices.
Release 7 2 3GPP TR 25.814 V7.1.0 (2006-09)
Keywords
UMTS, radio, packet mode,
layer 1
3GPP
Postal address
Internet
http://www.3gpp.org
Copyright Notification
Contents
Foreword ............................................................................................................................................................ 8
1 Scope ....................................................................................................................................................... 9
1.1 Rationale for RAN#30 decision on way forward for Evolved UTRA multiple access....................................... 9
2 References ............................................................................................................................................. 10
3 Definitions, symbols and abbreviations................................................................................................. 11
3.1 Definitions ............................................................................................................................................................... 11
3.2 Symbols ................................................................................................................................................................... 11
3.3 Abbreviations .......................................................................................................................................................... 11
4 Introduction ........................................................................................................................................... 12
5 Requirements ......................................................................................................................................... 13
6 General description of layer 1 ............................................................................................................... 13
6.1 Multiband operation................................................................................................................................................ 13
6.1.1 MC-WCDMA based proposal .......................................................................................................................... 13
6.2 Duplexing ................................................................................................................................................................ 13
6.2.1 TDD mode aspects for OFDMA and SC-FDMA .......................................................................................... 13
6.2.1.1 Approach 1 –Multiple fixed frame structures ........................................................................................... 14
6.2.1.1.1 For co-existence with LCR-TDD ......................................................................................................... 14
6.2.1.1.2 For co-existence with HCR-TDD ........................................................................................................ 15
6.2.1.2 Approach 2 –Generic frame structure ........................................................................................................ 16
6.2.2 Duplexing for the MC-WCDMA based proposal ........................................................................................... 19
6.2.2.1 Unpaired spectrum use cases ...................................................................................................................... 19
6.2.2.2 Time duplex operation ................................................................................................................................ 19
7 Downlink concepts ................................................................................................................................ 20
7.1 OFDMA (FDD / [TDD]) ........................................................................................................................................ 20
7.1.1 Basic transmission scheme ..................................................................................................................................... 20
7.1.1.1 Modulation scheme ..................................................................................................................................... 22
7.1.1.1.1 Basic modulation scheme ..................................................................................................................... 22
7.1.1.1.2 Enhanced modulation scheme .............................................................................................................. 22
7.1.1.2 Multiplexing including reference-signal structure .................................................................................... 26
7.1.1.2.1 Downlink data multiplexing ................................................................................................................. 26
7.1.1.2.2 Downlink reference-signal structure .................................................................................................... 27
7.1.1.2.3 Downlink L1/L2 Control Signaling ..................................................................................................... 28
7.1.1.3 Channel coding and physical channel mapping ........................................................................................ 30
7.1.1.4 MIMO and Transmit Diversity .................................................................................................................. 30
7.1.1.4.1 Transmit Diversity for Control Channel .............................................................................................. 30
7.1.1.4.2 Aspects for LTE MIMO ....................................................................................................................... 31
7.1.1.4.3 High level principles of MIMO for unicast traffic .............................................................................. 31
7.1.1.4.4 High level principles of MIMO for E-MBMS .................................................................................... 32
7.1.1.5 Downlink macro diversity .......................................................................................................................... 33
7.1.1.6 MBMS ......................................................................................................................................................... 33
7.1.2 Physical layer procedure ................................................................................................................................... 35
7.1.2.1 Scheduling ................................................................................................................................................. 35
7.1.2.2 Link adaptation.......................................................................................................................................... 35
7.1.2.3 HARQ ........................................................................................................................................................ 36
7.1.2.4 Cell search ................................................................................................................................................. 37
7.1.2.4.1 Purposes of the SCH, BCH, and reference symbols and information to be
detected in the cell search ..................................................................................................................... 37
7.1.2.4.2 Structure in time .................................................................................................................................... 38
7.1.2.4.3 Structure in frequency ........................................................................................................................... 40
7.1.2.4.4 Transmit diversity for SCH and BCH transmission............................................................................ 42
7.1.2.4.5 SCH signal structure ............................................................................................................................. 42
Release 7 4 3GPP TR 25.814 V7.1.0 (2006-09)
Foreword
This Technical Report has been produced by the 3rd Generation Partnership Project (3GPP).
The contents of the present document are subject to continuing work within the TSG and may change
following formal TSG approval. Should the TSG modify the contents of the present document, it will
be re-released by the TSG with an identifying change of release date and an increase in version number
as follows:
Version x.y.z
where:
y the second digit is incremented for all changes of substance, i.e. technical enhancements,
corrections, updates, etc.
z the third digit is incremented when editorial only changes have been incorporated in the
document.
Release 7 9 3GPP TR 25.814 V7.1.0 (2006-09)
1 Scope
This document is related to the technical report for physical layer aspect of the study item “Evolved
UTRA and UTRAN” [1]. The purpose of this TR is to help TSG RAN WG1 to define and describe the
potential physical layer evolution under consideration and compare the benefits of each evolution
techniques, along with the complexity evaluation of each technique.
This activity involves the Radio Access work area of the 3GPP studies and has impacts both on the
Mobile Equipment and Access Network of the 3GPP systems.
This document is intended to gather all information in order to compare the solutions and gains vs.
complexity, and draw a conclusion on way forward.
This document is a ‘living’ document, i.e. it is permanently updated and presented to TSG-RAN
meetings.
- When compared to the reference defined in TR 25.913, and based on the initial system-level
evaluations with 5 MHz allocation, the spectral efficiency improvements achievable with a
(CDMA-based) system according to an “evolutionary” approach and the spectral efficiency
improvements achievable with a new approach (e.g. OFDM-based) are both attractive.
- Using a CDMA based approach enables smoother migration from prior UTRA releases and
might offer more extensive physical layer reuse.
- On the other hand, a new Layer 1, with an inherent avoidance of a priori constraints in the air-
interface design, allows for a more free choice of design parameters, making it easier to fulfil
some of the E-UTRA targets e.g. latency requirements, finer minimum bandwidth granularity,
commonality between different duplex modes.
- UE receiver processing is somewhat simpler for an OFDMA-based air interface; the
attractiveness in terms of complexity increases with larger bandwidths and/or high order
MIMO configurations.
Both approaches to the 3GPP radio-access evolution have their advantages and disadvantages, very
much depending on the exact requirements.
On this basis, TSG-RAN #30 has decided that the Long-Term Evolution feasibility study will focus on
OFDMA based downlink and SC-FDMA based uplink. TSG-RAN #30 has also re-affirmed that
continued evolution of existing UTRA modes is an on-going necessary work activity within 3GPP.
Release 7 10 3GPP TR 25.814 V7.1.0 (2006-09)
2 References
The following documents contain provisions which, through reference in this text, constitute provisions
of the present document.
• References are either specific (identified by date of publication, edition number, version
number, etc.) or non-specific.
• For a non-specific reference, the latest version applies. In the case of a reference to a 3GPP
document (including a GSM document), a non-specific reference implicitly refers to the latest
version of that document in the same Release as the present document.
[1] 3GPP TD RP-040461: "Proposed Study Item on Evolved UTRA and UTRAN".
[2] 3GPP, TR25.848, “Physical Layer Aspects of UTRA High Speed Downlink Packet Access.”
[3] 3GPP, TR25.896, “Feasibility Study for Enhanced Uplink for UTRA FDD”
[4] 3GPP TR 25.913, “Requirements for evolved Universal Terrestrial Radio Access (UTRA) and
Universal Terrestrial Radio Access Network (UTRAN)”
[6] ETSI TR 101 112 (V3.1.0): “Universal Mobile Telecommunications System (UMTS); Selection
procedures for the choice of radio transmission technologies of the UMTS (UMTS 30.03 version
3.1.0).”
[7] 3GPP, TR25.996, “Spatial Channel Model for Multiple Input Multiple Output (MIMO)
[9] 3GPP, R1-040642, “Comparison of PAR and Cubic Metric for Power De-rating”, Motorola
[10] 3GPP, R4AH-05045, “UE transmit configuration and E-TFC Selection”, Motorola
[11] Nielsen et al, “Statistic of Measured Body Loss for Mobile Phones”, IEEE Trans. Antennas and
Propagation, VOL. 49, No.9, Pages 1351-3, September 2001.
[12] Kostanic et al, “Measurements of the Vehicle Penetration Loss Characteristics at 800MHz”, IEEE
VTC 1998, Volume 1, 18-21 May 1998 Page(s): 1-4 vol. 1.
[13 ] Davidson and Hill, “Measurement of Building Penetration into Medium Buildings at 900 and
1500 MHz, “IEEE VTC, VOL. 46, NO. 1, February 1997, Pages 161-168.
[14] Mohammed et al, “Characterization of Indoor Penetration Loss at ISM Band”, Environmental
Electromagnetics, 2003. CEEM 2003. Proceedings. 4-7 Nov. 2003, Page(s): 25-28.
[15] Zhang and Hwang, “Measurements of the Characteristics of Indoor Penetration Loss”, IEEE VTC
1994, 8-10 June 1994, Pages: 1741-1744 vol 3.
[16] 3GPP R1-061118 “E-UTRA physical layer framework for evaluation”, Vodafone, Cingular,
DoCoMo, Orange, Telecom Italia, T-Mobile, Ericsson, Qualcomm, Motorola, Nokia, Nortel, Samsung,
Siemens
[17]3GPP RP-060169, “Concept evaluation for evolved UTRA/UTRAN”, Cingular Wireless, CMCC,
NTT DoCoMo, O2, Orange, Telecom Italia, Telefonica, T-Mobile, Vodafone.
Release 7 11 3GPP TR 25.814 V7.1.0 (2006-09)
[18] J.H. Winters, “Optimum Combining on Digital Mobile Radio with Co-channel Interference”,
IEEE Journal on Selected Areas in Communication, Vlo. SAC-2, No.4, July 1984.
3.1 Definitions
For the purposes of the present document, the following terms and definitions apply.
3.2 Symbols
For the purposes of the present document, the following symbols apply:
<symbol> <Explanation>
3.3 Abbreviations
For the purposes of the present document, the following abbreviations apply:
ACK Acknowledgement
ASESS Adaptive Selection of the Surviving Symbol Replica Candidates
BCH Broadcast Channel
BS Base Station
CAZAC Constant Amplitude Zero Auto-Correlation
CDD Cyclic Delay Diversity
CP Cyclic Prefix
CQI Channel Quality Indicator
CRC Cyclic Redundancy Check
DCT Discrete Cosine Transform
DL Downlink
DRX Discontinuous Reception
DSCH Downlink Shared Channel
DTX Discontinuous Transmission
DUSP Switching point from downlink to uplink
E-DCH Enhanced Dedicated Channel
E-UTRA Evolved UTRA
E-UTRAN Evolved UTRAN
FBI Feedback Information
FDD Frequency Division Duplex
FDM Frequency Division Multiplexing
FEC Forward Error Correction
FFT Fast Fourier Transform
FSTD Frequency Switched Transmit Diversity
GERAN GSM EDGE Radio Access Network
HARQ Hybrid Automatic Repeat reQuest
HCR High Chip Rate
HSDPA High Speed Downlink Packet Access
IDFT Inverse Discrete Fourier Transform
IDMA Interleaved Division Multiple Access
IFFT Inverse Fast Fourier Transform
Release 7 12 3GPP TR 25.814 V7.1.0 (2006-09)
4 Introduction
At the 3GPP TSG RAN #26 meeting, the SI description on “Evolved UTRA and UTRAN” was
approved [1].
The justification of the study item was, that with enhancements such as HSDPA and Enhanced Uplink,
the 3GPP radio-access technology will be highly competitive for several years. However, to ensure
competitiveness in an even longer time frame, i.e. for the next 10 years and beyond, a long-term
evolution of the 3GPP radio-access technology needs to be considered.
Important parts of such a long-term evolution includes reduced latency, higher user data rates,
improved system capacity and coverage, and reduced cost for the operator. In order to achieve this, an
evolution of the radio interface as well as the radio network architecture should be considered.
Considering a desire for even higher data rates and also taking into account future additional 3G
spectrum allocations the long-term 3GPP evolution should include an evolution towards support for
wider transmission bandwidth than 5 MHz. At the same time, support for transmission bandwidths of
5MHz and less than 5MHz should be investigated in order to allow for more flexibility in whichever
frequency bands the system may be deployed.
Release 7 13 3GPP TR 25.814 V7.1.0 (2006-09)
5 Requirements
(Editor’s note: we refer the related requirement in TR25.913)
The multi-carrier approach is also well suited to simultaneous operation in multiple bands and over
discontinuous allocations in the same band. Multi-band operation is primarily limited by the UE RF
complexity and the laws of physics associated with the distance between the bands.
6.2 Duplexing
The E-UTRA air interface supports both frequency division duplex (FDD) and time division duplex
(TDD) modes of operation.
The downlink and uplink concepts described in sections 7.1, 9.1 and 9.2 are common to FDD and TDD
modes of operation unless otherwise stated (i.e. where specific properties/capabilities of FDD and TDD
duplex arrangements need to be taken into account).
Downlink synchronisation reference signals and system information are contained in each frame and
only occupy parts of the frame. If the synchronisation and system information signal structures are
common to TDD and FDD modes, then it may be possible to realise some benefits in terms of UE
complexity.
The traffic sub frame structure for TDD mode operation of E-UTRA is shown in Fig. 6.2.1-1. This
structure supports timing advance for cells of various sizes.
Release 7 14 3GPP TR 25.814 V7.1.0 (2006-09)
tFFS µs tFFS µs
E-UTRA, when operating in TDD mode-of-operation, may face additional interference scenarios,
compared to when operating in FDD mode of operation. More specifically, direct UE-to-UE and BS-to-
BS interference may occur both within one carrier and between neighbour carriers.
If E-UTRA operates in TDD mode, two approaches have been proposed to meet the requirement on co-
existence with current UTRA TDD according to TR 25.913. The possibility of adopting both the two
approaches can be considered.
Approach 1 is described in section 6.2.1.1, and approach 2 is described in section 6.2.1.2. Either
approach can be used depending on co-existence scenario.
In order to meet the latency requirement of TR25.913, it may be necessary to employ additional
switching points in E-UTRA TDD compared to UTRA TDD. Any interference problem created as a
result between the E-UTRA and the legacy carrier would need to be solved.
A guard period is required at a DL/UL or UL/DL switching point. Each traffic time slot should contain
a small idle period (Timeslot Interval) which can be used for switching guard period from UL to DL
traffic time slot. For the DL/UL switching point, if there is only one switching point between DL and
UL traffic time slot in a E-UTRA TDD radio traffic time slot structure, a special guard period will be
kept between the special downlink timeslot DwPTS and special uplink timeslot UpPTS.
Release 7 15 3GPP TR 25.814 V7.1.0 (2006-09)
Radio Sub-frame
(5ms)
DUSP UDSP
Figure 6.2.1.1-1 Frame Structure of one pair of switching points between DL and UL traffic timeslot
Assuming that a 10 ms frame is divided into 2 equally sized 5 ms radio sub-frames, one radio sub-
frame consists of seven traffic time slots (TS0~TS6). The synchronization and guard period is between
TS0 and TS1, whose duration is 0.275ms including DwPTS, GP and UpPTS. The TTI of E-UTRA
TDD can be 0.675ms, the same as duration of one traffic time slot.
Note: as mentioned in Section 6.2.1, special considerations need to be taken with regards to the E-
UTRA frame structure when operating in TDD mode of operation in order to satisfy the E-UTRA
requirement on allowing for spectrum co-existence with current 3GPP TDD standards.
The minimum TTI for uplink transmission is equal to the uplink traffic timeslot duration (0.675ms).
A guard period is required at a DL/UL or UL/DL switching point. Each traffic time slot should contain
a small idle period (Timeslot Interval) which can be used for switching guard period from UL to DL
traffic time slot.
To facilitate full alignment of the uplink and downlink transmission periods on an E-UTRA carrier
with those of a 3.84Mcps TDD carrier for all possible UL/DL timeslot slot configurations, an E-UTRA
sub-frame duration of 0.01/(15n) seconds is required, where n = {1,2,3,…}.
However, many common 3.84Mcps TDD deployments use a single switching point per radio frame.
This, taken in conjunction with the fact that 3 HCR timeslots are of equal duration to 4x0.5ms sub-
frames for E-UTRA (see for example section 7.1.1, 9.1.1, 9.2.1), provides for further E-UTRA/UTRA
TDD frame alignment possibilities when an E-UTRA sub-frame duration of 0.5ms is selected.
1. Via an E-UTRA sub-frame duration of 0.01/(15n) seconds. Each E-UTRA sub-frame within a
10ms radio frame may be configured as uplink or downlink
2. Via the use of the generic frame structure of 6.2.1.2 with 0.5ms sub-frame duration.
Release 7 16 3GPP TR 25.814 V7.1.0 (2006-09)
For reasons of commonality with the paired LTE mode, option 2 above is preferred.
HCR-TDD coexistence
The HCR-TDD timeslot duration THCR-TDD is related to the E-UTRA sub-frame duration TE-UTRA of
sections 7.1 and 9.1 / 9.2 according to the relationship:
Hence, the E-UTRA uplink and downlink may be aligned with the HCR-TDD uplink and downlink
provided that the HCR-TDD UL:DL timeslot split is of the form 3 × n : 3 × (5 − n ) , where n is an integer.
In this case, the E-UTRA UL:DL split is 4 × n : 4 × (5 − n ) . An example alignment of the HCR-TDD
frame to the E-UTRA frame is shown in Figure 6.2.1.2-1 for a 6:9 UL:DL timeslot split. The flexible
frame structure of HCR-TDD allows existing HCR-TDD deployments to be migrated to a
3 × n : 3 × (5 − n ) timeslot split in readiness for a future E-UTRA deployment in an adjacent carrier. As
per the LCR-TDD case, idle symbols / sub-frames may be inserted into the E-UTRA carrier when the
HCR-TDD carrier does not utilise a 3 × n : 3 × (5 − n ) UL:DL split.
10ms
HCR-TDD 0.667ms
E-UTRA
0.5ms
2ms
Figure 6.2.1.2-1 - E-UTRA / HCR-TDD co-existence example for 6:9 UL: DL timeslot split
LCR-TDD coexistence
Coexistence between LCR-TDD and E-UTRA may be facilitated by inserting either idle symbols
within the E-UTRA frame (these are required for the purposes of timing advance in any case) or idle
sub-frames (which may either be inserted by the E-UTRAN scheduler dynamically or their existence
may be signalled on the broadcast control channel). Applying a delay or frame offset between the LCR-
TDD frame and the E-UTRA frame may allow the time allocated to idle symbols / sub-frames to be
minimised. Figure 6.2.1.2-2 shows how coexistence between LCR-TDD with a 3:3 UL:DL traffic
timeslot split and E-UTRA (operating in a TDD mode with the numerology of sections 7.1 and 9.1 /
9.2) can be facilitated. To increase spectral efficiency, the idle sub-frame of Figure 6.2.1.2-2 could be
replaced by 4 data symbols followed by 3 idle symbols. Similarly, Figure 6.2.1.2-3 shows facilitation
of coexistence of between LCR-TDD with a 2:4 UL:DL traffic timeslot split and E-UTRA. Note that
the idle sub-frame shown in Figure 6.2.1.2-2 is only required for an adjacent E-UTRA carrier. For a
non-adjacent E-UTRA carrier, the idle E-UTRA sub-frame can be replaced by either a downlink E-
UTRA sub-frame or an uplink E-UTRA sub-frame.
Release 7 17 3GPP TR 25.814 V7.1.0 (2006-09)
10ms
5ms 5ms
LCR-TDD
E-UTRA
5ms
750
825
1625
2300
2975
3650
4325
5000
675
950
0
LCR-TDD
DL UL UL UL DL DL DL
DwPTS UpPTS
GP idle SC-FDMA block /
E-UTRA idle sub-frame OFDM symbol
DL UL UL UL UL DL DL DL DL
0
500
1000
1500
2000
2500
3000
3500
4000
4500
5000
time / µs
Figure 6.2.1.2-2 - E-UTRA / LCR-TDD co-existence example for 3:3 UL: DL timeslot split
Release 7 18 3GPP TR 25.814 V7.1.0 (2006-09)
10ms
5ms 5ms
LCR-TDD
E-UTRA
5ms
750
825
1625
2300
2975
3650
4325
5000
LCR-TDD
675
950
0
DL UL UL DL DL DL DL
DwPTS UpPTS
GP idle OFDM symbol idle OFDM symbol
E-UTRA
DL DL UL UL UL UL DL DL DL DL DL
325
825
1325
1825
2325
2825
3325
3825
4325
4825
time / µs
Figure 6.2.1.2-3 - E-UTRA / LCR-TDD co-existence example for 2:4 UL: DL timeslot split
In the case where there are no coexistence issues, this frame structure degenerates to the frame structure
of Figure 6.2.1.2-4. This figure is for illustrative purposes only. Idle symbols are required only at
DL/UL switching points. The idle period, required in the Node B at UL/DL switching points, is created
by timing advance means.
Release 7 19 3GPP TR 25.814 V7.1.0 (2006-09)
10 ms Radio Frame
sub-frames (0.5ms)
OFDMA DL SC-FDMA / OFDM UL
IP sub-frame
e.g. 4 standard
e.g. 5 standard sub-frames sub-frames
sub-channel index
sub-frame sub-frame
7 symbols 6 symbols + IP
Tsf Tsf
• Use the unpaired band allocation for the downlink only transmission of the physical channels
in support of E-MBMS; this enables standalone usage of the unpaired spectrum, possibly in
time multiplexed manner with legacy UTRA-TDD deployments.
• Use the unpaired band allocation as additional downlink or uplink carriers to support
asymmetric multi-carrier configurations.
It would be possible to introduce these new modes of operation as part of the MC-WCDMA based
proposal. However, this would require significant changes to the original UTRA-FDD channel
structure, timing and procedures and would therefore not benefit in the same way from the UTRA-FDD
learning curve.
A more natural approach which is consistent with the underlying backward compatibility philosophy is
therefore to consider the existing UTRA-TDD modes and their multi-carrier evolution for standalone
time duplex operation in unpaired spectrum allocations.
7 Downlink concepts
Three basic concepts are proposed in downlink:
2. MC-WCDMA (FDD)
3. MC-TD-SCDMA (TDD)
†† This is the assumption for the baseline proposal. Somewhat more carriers may be possible to occupy in case of
the wider bandwidth
Release 7 21 3GPP TR 25.814 V7.1.0 (2006-09)
*: {(x1/y1) × n1, (x2/y2) × n2} means (x1/y1) for n1 OFDM symbols and (x2/y2) for n2 OFDM
symbols
The sub-frame duration corresponds to the minimum downlink TTI. The possibility to concatenate
multiple sub-frames into longer TTIs, e.g. for improved support for lower data rates and QoS
optimization, should be considered. In this case, the TTI can either be a semi-static or dynamic
transport channel attribute. In case of a semi-static TTI, the TTI is set through higher layer signalling.
In case of a dynamic TTI, the number of sub-frames concatenated can be dynamically varied for at
least the initial transmission and possibly for retransmissions. It is to be determined to what extent a
dynamic TTI can reduce higher layer protocol overhead (e.g. MAC, RLC), L1 overhead (e.g. CRC),
and ACK/NACK feedback, as well as reducing latency by reducing segmentation of IP packets. It is
initially assumed that the Network (e.g. Node-B) would signal the TTI, either explicitly (e.g. with L1
bits) or implicitly (e.g. by indicating modulation and coding rate and transport block size). The
interaction between dynamic TTI, signaling errors, HARQ procedure (time synchronous vs.
asynchronous including adaptive or non-adaptive characteristics) and UE complexity needs to be
investigated.
Note that the sub-carrier spacing is constant regardless of the transmission bandwidth. To allow for
operation in differently sized spectrum allocations, the transmission bandwidth is instead varied by
varying the number of OFDM sub-carriers. The necessity for supporting an additional longer cyclic-
prefix duration, see Table 7.1.1-1, is under consideration. The longer cyclic prefix should then target
multi-cell broadcast and very-large-cell scenarios.
The mapping and indexing of N available physical channel symbols (sub-carriers) in one OFDM
symbol in RF spectrum should be done as illustrated in figure below
For E-UTRA TDD, the frame structure corresponding to Table 7.1.1-1 is supported. In addition, a
second frame structure is also supported with the intention of providing co-existence with LCR UTRA
TDD. The sampling frequency, FFT size, sub-carrier spacing, and number of occupied sub-carriers is
the same as for Table 7.1.1-1. However, with this alternative frame structure, a 10 ms radio frame is
divided into 2 equally sized 5 ms sub-frames1 (of which a subset is allocated for downlink
transmission), one sub-frame consists of seven traffic time slots (TS0~TS6) and three special time
slots, and one example is shown in Figure 6.2.1.1-1. The synchronization and guard period is between
TS0 and TS1, whose duration is 0.275ms. Each time slot should contain a small idle period (Timeslot
Interval) which can be used for switching guard period from UL to DL time slots. The basic
transmission parameters for this alternative frame structure are specified in Table 7.1.1-2 below.
1 Note that the term “sub-frame” is, in this case, aligned to the LCR UTRA TDD terminology and not to the terminology
currently used for E-UTRA. The E-UTRA term “sub-frame” corresponds to the term “time slot” used here.
Release 7 22 3GPP TR 25.814 V7.1.0 (2006-09)
Table 7.1.1-2 - Parameters for downlink transmission scheme (alternative TDD frame structure)
†† This is the assumption for the baseline proposal. Somewhat more carriers may be possible to occupy in case of
the wider bandwidth
Contrary to conventional OFDM modulation, OFDM/OQAM modulation does not require a guard
interval (also called cyclic prefix). For this purpose, the prototype function modulating each sub-carrier
must be very well accurately localized in the time domain, to limit the inter-symbol interference for
transmissions over multipaths channels.
This prototype function can be also accurately localized in the frequency domain, to limit the inter-
carrier interferences (due to Doppler effects, phase noise…). This function must also guarantee
orthogonality between sub-carriers both in time and frequency domains.
It is mathematically proven that when using complex valued symbols, the prototype functions
guaranteeing perfect orthogonality at critical sampling rate can not be well localised both in time and
frequency. For instance the unity function used in conventional OFDM has weak frequency localisation
properties and obliges using a cyclic prefix between the symbols to limit inter-symbol interference.
To let the use of accurately localised functions in the time-frequency domain OFDM/OQAM
introduced a time offset between the real part and the imaginary part of the symbols. Orthogonality is
then guaranteed only over real values. The corresponding multi-carrier modulation is an
OFDM/OQAM. The OFDM/OQAM transmitted signal is expressed as
Release 7 23 3GPP TR 25.814 V7.1.0 (2006-09)
M −1
s(t ) = ∑ ∑ a m, n i m + n e 2iπmν 0t g (t − nτ 0 ) (7.1.1.1.2-1)
n m=0
LOOOMOOON
g m , n (t )
where am,n denotes the real valued information value (can be the real part or the imaginary part of the
Offset complex QAM symbol) sent on the mth sub-carrier at the nth symbol, M is the number of sub-
carriers, ν0 is the inter-carrier spacing, it is the same of the classical OFDM system. τ0 is the
OFDM/OQAM symbol duration, it is equal to Tu/2 (Tu is the OFDM symbol duration), and g is the
prototype function.
It is important to notice that OFDM/OQAM symbol rate is twice the classical OFDM symbol rate
without cyclic prefix (τ0 =N/2), meanwhile, since the modulation used is a real one, the information
amount sent by an OFDM/OQAM symbol is half the information amount sent by an OFDM symbol.
Figure 7.1.1.1.2-1 depicts the signal generation chain of an OFDM/OQAM signal.
The modulator generates N real valued symbols, each τ0 where τ0 = Tu / 2. The real valued symbols are
then dephased, they are multiplied by im+n before the IFFT as it is noted in (7.1.1.1.2-1). Figure
7.1.1.1.2-2 shows the time-frequency localisation of the transmitted symbols both for conventional
OFDM using complex valued QAM symbols and for OFDM/OQAM.
ν0
τ0
2×τ0 = Tu
Figure 7.1.1.1.2-2: OFDM/OQAM time and frequency lattices (compared to conventional OFDM
w/o guard interval)
The main difference of OFDM/OQAM over conventional OFDM signal generation stays in the
filtering by the prototype function g after the IFFT, instead of the cyclic prefix addition.
Release 7 24 3GPP TR 25.814 V7.1.0 (2006-09)
Thanks to the Inverse Fourier Transform, the prototype function g can be implemented in its polyphase
form, which reduces strongly the complexity of the filtering. Moreover the density 2 induces some
more simplifications in the polyphase implementation. Figure 7.1.1.1.2-3 shows a possible polyphase
implementation of both an OFDM/OQAM modulator and demodulator (Gi are the polyphase
components of the prototype filter).
One candidate for OFDM/OQAM filter (g) is IOTA (Isotropic Orthogonal Transform Algorithm)
prototype obtained by orthogonalizing the Gaussian function in both time and frequency domains
according to Schmidt method. See Figure 7.1.1.1.2-4.
IOTA Fourier Transform
IOTA
1.2 1.2
1 1
0.8 0.8
Amplitude
Amplitude
0.6 0.6
0.4 0.4
0.2 0.2
0 0
-0.2 -0.2
-0.4 -0.4
- 3τ 0 -2*τ 0 −τ 0 0 τ0 2*τ 0 3τ 0 - 3ν0 -2*ν0 −ν0 0 ν0 2*ν0 3ν0
Figure 7.1.1.1.2-4: Temporal and spectral representation of the IOTA prototype filter
Another particularity of IOTA is the spectrum of the generated signal. Thanks to its good frequency
localisation, the resulting spectrum is steeper than for conventional OFDM. Figure 7.1.1.1.2-5 depicts
the resulting spectra of the signals generated by both rectangular filter used in classical OFDM system
and IOTA function used for OFDM/OQAM system. FFT length is 512 and 300 sub-carriers are
modulated (parameters corresponding to the 5 MHz case)
Release 7 25 3GPP TR 25.814 V7.1.0 (2006-09)
The OFDM/OQAM transmission scheme is very similar to the conventional OFDM scheme listed in
Table 7.1.1-1, with a sub-carrier spacing ∆f = 15 kHz. Assuming that a 10 ms radio frame is divided
into 20 equally sized sub-frames, this parameter set implies a sub-frame duration Tsub-frame = 0.5 ms. As
for conventional OFDM it may be noted that numerology specified below are for evaluation purpose
only. All remarks regarding the support of concatenated TTI remain relevant.
CP length 0
†† This is the assumption for the baseline proposal. Somewhat more carriers may be possible to occupy in case of
the wider bandwidth
*
: In OFDM/OQAM the symbol rate is twice higher than for conventional OFDM (if no CP was included) and the
amount of information transmitted per OFDM/OQAM symbol is half the amount transmitted by 1 conventional
OFDM symbol (see section 7.1.1.1.2.1 for more details)
Release 7 26 3GPP TR 25.814 V7.1.0 (2006-09)
Table 7.1.1.2.1-1 Physical resource block bandwidth and number of physical resource blocks
dependent on bandwidth.
Using other values such as, e.g. M=15 or M=12 or M=10 or M equal to other values can be considered
based on the outcome of the interference coordination study.
The frequency and time allocations to map information for a certain UE to resource blocks is
determined by the Node B scheduler and may e.g. depend on the frequency-selective CQI (channel-
quality indicator) reported by the UE to the Node B, see Section 7.1.2.1 (time/frequency-domain
channel-dependent scheduling). The channel-coding rate and the modulation scheme (possibly different
for different resource blocks) are also determined by the Node B scheduler and may also depend on the
reported CQI (time/frequency-domain link adaptation).
Distributed VRBs are mapped onto the PRBs in a distributed manner. Localized VRBs are mapped
onto the PRBs in a localized manner. The exact rules for mapping VRBs to PRBs are FFS.
The multiplexing of localized and distributed transmissions within one subframe is accomplished by
FDM.
As a result of mapping VRBs to PRBs, the transmit bandwidth is structured into a combination of
localized and distributed transmissions. Whether this structuring is allowed to vary in a semi-static or
dynamic (i.e. per subframe) way is FFS. The UE can be assigned multiple VRBs by the scheduler. The
information required by the UE to correctly identify its resource allocation must be made available to
the UE by the scheduler. The number of signalling bits required to support the multiplexing of
localized and distributed transmissions should be optimized.
Release 7 27 3GPP TR 25.814 V7.1.0 (2006-09)
Details of the multiplexing of lower-layer control signaling is currently TBD but may be based on time,
frequency, and/or code multiplexing.
- Downlink-channel-quality measurements
- Downlink channel estimation for coherent demodulation/detection at the UE
- Cell search and initial acquisition
The basic downlink reference-signal structure, consisting of known reference symbols, is illustrated in
Figure 7.1.1.2.2-1.
Reference symbols (a.k.a. ”First reference symbols”) are located in the first OFDM symbol of every
sub-frame assigned for downlink transmission. This is valid for both FDD and TDD as well as for both
long and short CP.
Additional reference symbols (a.k.a. ”Second reference symbols”) are located in the third last OFDM
symbol of every sub-frame assigned for downlink transmission. This is the baseline for both FDD and
TDD as well as for both long and short CP. However, it should be evaluated if, for FDD, the second
reference symbols are needed.
Frequency domain
D R1 D D D D D R1 D D D D D R1 D D D D D
D D D D D D D D D D D D D D D D D D D
D D D D D D D D D D D D D D D D D D D
0.5 ms
D D D D D D D D D D D D D D D D D D D
D D D D R2 D D D D D R2 D D D D D R2 D D
D D D D D D D D D D D D D D D D D D D
D D D D D D D D D D D D D D D D D D D
The spacing (in the frequency domain) between reference symbols of the same OFDM symbol and
antennas is M = 6 sub-carriers [also other values of M could be studied]. The first and second reference
symbols are staggered in the frequency domain as illustrated in Figure 7.1.1.2.2-1 above
The current assumption is that the position (in the frequency domain) of the reference symbols may
vary from sub-frame to sub-frame and between cells. However, this assumption may be reconsidered if
it is in conflict with any future conclusions regarding the E-UTRA cell-search procedure.
In the case that Layer 1 downlink control signaling (more specifically signaling or part of the signaling
related to downlink and uplink scheduling) is located at the beginning of the corresponding sub-frame
Release 7 28 3GPP TR 25.814 V7.1.0 (2006-09)
(still TBD if this will be the case or if the Layer 1 signaling is to be spread over the sub-frame), it is
currently assumed that demodulation of this information could be carried out without using the second
reference symbols of the corresponding sub-frame (however, second reference symbols of previous
sub-frames may be used if available).
- To support transmission using multiple TX antennas within one cell (up to a maximum of 4
orthogonal reference signals should be supported to enable higher-order downlink MIMO)
within one cell/beam. Note that, for TDD, orthogonal reference signals may not be needed
between TX antennas of the same Node B if the multiple TX antennas are used for downlink
dynamic beam forming.
- To allow for orthogonal reference signals between sectors and fixed beams of the same Node
B.
Orthogonality between reference signals of different TX antennas of the same cell/beam is created by
means of FDM. This implies that the reference-signal structure of Figure 7.1.1.2.2-1, with different
antenna-specific frequency shifts, is valid for each antenna. CDM should be evaluated as an alternative.
In case of orthogonality between reference signals of different cells/beams belonging to the same Node
B, the orthogonality is created in the code domain, i.e. the (frequency domain) sequence of reference
symbols are multiplied by mutually orthogonal patterns. [The assumption regarding CDM-based
orthogonality between reference signals of cells/beams of the same Node B is to be confirmed by means
of system-level evaluation.]
Possible transmission of additional UE-specific downlink reference symbols are to be considered for
dynamic beam forming or MIMO.
Cell common reference signal is the baseline reference signal structure for multi-cell MBMS
transmissions. Support of cell-specific reference signals together with group scrambling is FFS. A
decision on multi-cell MBMS reference signal structure will be made based on throughputs offered by
each candidate scheme under evaluation conditions to be agreed to.
Downlink scheduling information is used to inform the UE how to process the downlink data
transmission. The information signalled to a UE scheduled to receive user data is summarized in Table
7.1.2.3.1-1.
The possible downlink time/frequency location(s) for category 1 information is known to the UE a
priori.
The category 3 information is transmitted for every TTI of data to the scheduled user(s).
Release 7 29 3GPP TR 25.814 V7.1.0 (2006-09)
is adopted
Cat. 3
Note: It is FFS whether asynchronous or synchronous hybrid ARQ operation will be adopted.
Note: In case of multi-layer transmission to a UE, multiple instances of (parts of) category 2 and
category 3 information may be required.
Note: It is FFS whether information about multi-layer transmission is included in ‘resource assignment’
or ‘multi-antenna related information’.
Uplink scheduling grants are used to assign resources to UEs for uplink data transmission. The
information signalled to a UE receiving an uplink scheduling grant is summarized in Table 7.1.1.2.3.2-
1. The modulation and coding scheme to use for uplink transmission is implicitly given by the resource
assignment and the transport format.
Release 7 30 3GPP TR 25.814 V7.1.0 (2006-09)
Note: It is FFS whether the transport format the UE uses is mandated by the Node B or controlled by
the UE.
7.1.1.2.3.3 ACK/NAK
The hybrid ARQ feedback in response to uplink data transmission consists of a single ACK/NAK bit.
To achieve high processing gain, repetition coding can be used as a complement to FEC.
the downlink control channels. Transmit diversity schemes vary with respect to their complexity and
ability to support a variable number of transmit antennas. Therefore, the simplicity and scalability of
transmit diversity schemes should be compared as well as their performance gains. Open loop block
code-based transmit diversity (STBC or SFBC), cyclic delay diversity, time (or frequency) switched
transmit diversity, pre-coded transmission (or adaptive beam forming), and combined space-time (or
space-frequency) block code/cyclic delay diversity techniques should be considered. The cyclic delay
diversity is assumed for L1/L2 control channels as the baseline for evaluations with two or four
transmit antennas at the cell site.
- Realistic assumptions have to be taken into account when comparing different MIMO
concepts, such as feedback errors and delays, needed multi-antenna reference signal overhead
and its effect on performance, complexity and signalling requirements etc. The resulting
reference signal and signalling overheads in both uplink and downlink have to be justified by
the shown improvements
Modes of operation of multiple transmit antennas at the cell site (denoted as MIMO mode) are spatial
multiplexing, beamforming, and single-stream transmit diversity mode(s). The MIMO mode is
restricted by the UE capability, e.g. number of receive antennas, and is determined taking into account
the slow channel variation. The MIMO mode should be adapted slowly (e.g. only at the beginning of
communication or every several 100 msec), in order to reduce the required control signalling (including
feedback) required to support the MIMO mode adaptation
- Multiple codewords (including single codeword as a special case) that use the same time-
frequency(-code) resource and are independently channel-coded with independent CRC
should be investigated. Possible values for the maximum number of codewords per resource
block transmitted by the Node B are 1, 2, , or 4. The maximum number of codewords that can
be received by a UE is either 1, 2, , or 4 and should be fixed depending only on the UE
capability. The number of code words per UE may be restricted to, e.g. 2 to reduce the control
signalling overhead.
- Link adaptation as a means to track fast channel variation is applied independently for each
codeword if multiple codewords are transmitted to a UE with control interval of 0.5ms – a few
ms.
- In addition to the SDM, the spatial division multiplexing of the modulation symbol streams
for different UEs using the same time-frequency(-code) resource is supported for evaluation,
which may be denoted as spatial division multiple access (SDMA) or multi-user (MU)-
MIMO. Note that the SDM is a special case of the SDMA. MU-MIMO is supported only with
the pre-coding. To avoid very complex UE receiver, cancellation of other UE’s interfering
signal is not assumed to be a mandatory feature. However, ways to aid the inter-UE
interference cancellation by providing necessary information are not ruled out. The application
of SU- or MU-MIMO to a UE is determined by Node B either dynamic or semi-static manner.
Release 7 32 3GPP TR 25.814 V7.1.0 (2006-09)
- Use of pre-coding as a means to convert the antenna domain MIMO signal processing into the
beam domain processing should be investigated. It is FFS whether the precoding should be
unitary or non-unitary. Because precoding might require less complex (linear) receivers to
achieve a given level of performance, both the receiver complexity requirements and the
performance of MIMO algorithms should be considered. Any additional feedback required
for precoding and any additional related computational complexity in the UE should also be
taken into account. Both codebook based pre-coding and non-codebook based pre-coding
should be considered. Codebook based methods select pre-coding vector(s) from a
codebook(s) to reduce the signalling overhead. The size of the codebooks should be
minimized and the codebooks should be static. Irrespective whether the codebook is used or
not, the amount of feedback should be minimized. The update interval of the selected pre-
coding vector(s) should be sufficiently short to track instantaneous channel variation (0.5 ms
or longer). Adaptive beam forming is an alternative method to achieve SDMA.
- Rank adaptation (and/or the antenna subset selection), of which exact implementation is FFS,
is supported for evaluation as a means to prevent possible performance loss from using higher
number of MIMO layers than can be supported by the channel condition. The number of
codewords transmitted to a UE is controlled through rank adaptation if the UE supports
multiple codewords, and is constrained by the number of layers.
- A transmit diversity technique can be applied when the rank is determined to be one. Possible
candidates for the transmit diversity mode are
- The cyclic delay diversity is the baseline for simulation of the transmit diversity mode for
unicast data channels. It is noted that in the end, one transmit diversity scheme should be
selected for the unicast data channel to reduce the number of unnecessary options.
- Followings are identified as candidates for the UE feedback information but not limited to:
o MIMO channel state information, which may be used by the Node B to determine
MIMO processing consisting of e.g., selection of the rank (and/or the antenna
subset), and/or the pre-coding, etc.
o If, for TDD operation, uplink reference signals should be transmitted to provide
channel state information to support downlink MIMO transmission.
o Channel quality indicator (CQI), which may be used by the Node B to decide a MCS
level(s). In addition to the CQI, another form of feedback signalling, which may be
similar to the feedback information (FBI) as defined from Rel-99, should be
considered as a candidate feedback signalling.
o If multiple operation modes are defined, e.g. MIMO and open loop transmit
diversity,, and an open loop MIMO are supported, it may be needed for a UE to
inform the Node B of the indication of (preferred) operation mode.
- in E-MBMS, there will be a single or multiple transmitting Node B’s and multiple receiving
UEs
In the absence of any feedback from the UEs in E-MBMS, the potential candidates for MIMO are
either an open-loop transmit diversity scheme, an open-loop spatial multiplexing approach or a hybrid
combination of both. Any form of additional transmit diversity is not expected to bring any significant
benefit if the number of SFN cells is large enough, because E-MBMS already enjoys from frequency-
diversity due to delayed signals received from multiple cells. However, E-MBMS service becomes
bandwidth limited in an SFN operation and therefore spatial-multiplexing techniques become
attractive. Moreover, the received signal from multiple cells sees increasing decorrelation in an E-
MBMS environment which also favors spatial-multiplexing.
Both single code word and multi-code word spatial multiplexing schemes are considered in the study
item phase. In case of multi-code word spatial multiplexing, dynamic adaptation of modulation and
coding etc. for each code word is not possible due to absence of channel quality feedback. However,
different code words can potentially use different modulation and coding and/or power offsets etc. in a
semi-static fashion in order to enable efficient interference cancellation at the UE receiver. Since the
baseline UE has only two antennas, the number of broadcast codewords are limited to two. E-MBMS
for UE’s limited to single codeword reception capability should be further considered. E-MBMS
signals from Node B's with more than two transmit antennas should be transparent to the UE.
- Alternative #1
The same reference signal pattern is used for the transmissions from the two cells. In this case,
from a UE point-of-view the multi-cell transmission will be identical to a single-cell
transmission using a different antenna pattern,
- Alternative #2
Different reference signal patterns are used for the transmission from the two cells.
The impact of the two alternatives on the reference signal design and overhead, if any, needs further
investigations.
For multi-cell broadcast, soft combining of radio links should be supported, assuming a sufficient
degree of inter-Node-B synchronization, at least among a sub-set of Node B’s. Mechanisms should be
supported in E-UTRAN that allow the network to adapt which cells are in an SFN that will transmit an
MBMS service that may be soft combined.
7.1.1.6 MBMS
MBMS transmissions may be performed in the following two ways:
- Multi-cell transmissions
- Single-cell transmissions
In case of single-cell transmission, the MBMS traffic channel (MTCH) can be mapped to the DL
shared data channel (DL-SCH). In case of multi-cell transmissions, the MTCH may be mapped to
another transport channel type.
Release 7 34 3GPP TR 25.814 V7.1.0 (2006-09)
In case of multi-cell transmission the cells and content are synchronised to enable for the terminal to
combine the energy from multiple transmissions without additional receiver complexity. Tight inter-
cell synchronization, in the order of substantially less than the cyclic prefix, is assumed in order for the
UE to be able to combine multi-cell MBMS transmissions. Also, in case of cell-common reference
signals, the reference symbol design needs to be such that those reference symbols needed for MBMS
reception are identical in all cells to be considered for combining.
One dimension of the MBMS design is whether the MBMS transmission is actually sharing the same
carrier with unicast traffic or not. For instance, for mobile TV, MBMS data can be sent on a separate
carrier not carrying anything other than broadcast/MBMS related information. In that case, the MBMS
design shall be able to efficiently benefit from such a situation.
It is desirable to have MBMS transmissions at high instantaneous data rates so that low transmission
duty cycle per MBMS “channel” (source content) enables low power consumption for MBMS capable
UEs. This may impose some requirements in the multiplexing of different multicast source contents at
the physical layer. For the case of multiplexing unicast and multicast traffic within the same carrier,
enabling low UE power consumption should be considered in the evaluation of the unicast/multicast
multiplexing schemes depicted below.
In systems with unicast and multicast traffic multiplexed within the same carrier, FDM/TDM
transmission of unicast and multicast traffic is considered. TDM, i.e. multicast traffic being the only
traffic being transmitted in a given sub-frame, is a special case of FDM multiplexing (multicast traffic
and unicast traffic multiplexed in the the frequency domain within the same sub-frame),and is not
precluded, especially for system bandwidths lower than or equal to minimum UE RF capability. FDM
approach is needed to support system bandwidths larger than minimum UE RF capability. The MBMS
control information shall be designed to support both types of multiplexing and the actual configuration
will be deployment specific (based on e.g., system bandwidth). TDM multiplexing of different MBMS
streams is supported to minimize the MBMS reception time at the UE.
When there is no unicast data traffic in sub-frames carrying MBMS, it is to be considered if all the
reference signals of Figure 7.1.1.2.2-1 need to be transmitted. In case that not all the reference signals
are transmitted in certain sub frames, the impact on L1/L2 control, CQI and synchronization
performance needs to be carefully evaluated.
The associated L1/L2 control channel for MTCH may be transmitted less frequently than the associated
L1/L2 control channel for DL-SCH. At least the cell specific related MBMS control information is
transmitted in a cell-specific manner to allow multiplexing of single-cell and multi-cell MBMS traffic.
The L1 coding and modulation chain for MBMS transmissions is the same as unicast transmissions as a
baseline.
Release 7 35 3GPP TR 25.814 V7.1.0 (2006-09)
7.1.2.1 Scheduling
The Node B scheduler (for unicast transmission) dynamically controls which time/frequency resources
are allocated to a certain user at a given time. Downlink control signaling informs UE(s) what
resources and respective transmission formats have been allocated. The scheduler can instantaneously
choose the best multiplexing strategy from the available methods; e.g. frequency localized or frequency
distributed transmission. The flexibility in selecting resource blocks and multiplexing users (7.1.1.2)
will influence the available scheduling performance. Scheduling is tightly integrated with link
adaptation (7.1.2.2) and HARQ (7.1.2.3). The decision of which user transmissions to multiplex within
a given sub-frame may for example be based on
The use of power and modulation adaptation per resource block is FFS.
CRC attachment
Channel coding
HARQ functionality
including adaptive
coding rate
Physical channel
segmentation Number of assigned
(resource block mapping) resource blocks
Adaptive modulation
(common modulation is selected)
7.1.2.3 HARQ
Downlink hybrid ARQ (HARQ) should be based on Incremental Redundancy. Note that Chase
Combining is a special case of Incremental Redundancy and is thus implicitly supported as well.
- Synchronous HARQ implies that (re)transmissions for a certain HARQ process are restricted to
occur at known time instants. No explicit signaling of the HARQ process number is required as
the process number can be derived from, e.g., the sub-frame number.
- Asynchronous HARQ implies that (re)transmissions for a certain HARQ process may occur at
any time. Explicit signaling of the HARQ process number is therefore required.
In principle, synchronous operation with an arbitrary number of simultaneous active processes at a time
instant could be envisioned. In this case, additional signaling may be required. Asynchronous operation
already supports an arbitrary number of simultaneous active processes at a time instant. Furthermore,
note that, in a synchronous scheme, the transmitter may choose not to utilize all possible retransmission
instants, e.g., to support pre-emption. This may require additional signaling.
The various forms of HARQ schemes are further classified as adaptive or non-adaptive in terms of
transmission attributes, e.g., the Resource Block (RB) allocation, Modulation and transport block size,
and duration of the retransmission. Control channel requirements are described for each case.
- Adaptive implies the transmitter may change some or all of the transmission attributes used in
each retransmission as compared to the initial transmissions (e.g. due to changes in the radio
conditions). Hence, the associated control information needs to be transmitted with the
retransmission. The changes considered are:
o Modulation
o Resource Block allocation
o Duration of transmission
- Non-Adaptive implies that changes, if any, in the transmission attributes for the
retransmissions, are known to both the transmitter and receiver at the time of the initial
transmission. Hence, the associated control information need not be transmitted for the
retransmission.
With those definitions, the HS-DSCH in WCDMA uses an adaptive, asynchronous HARQ scheme,
while E-DCH in WCDMA uses a synchronous, non-adaptive HARQ scheme.
The capability of adaptively being able to change the packet format (i.e., adaptive IR) and the
transmission timing (i.e., asynchronous IR) yields an adaptive, asynchronous IR based HARQ
operation. Such a scheme has the potential of optimally allocating the retransmission resources in a
time varying channel. For each HARQ retransmission, control information about the packet format
needs to be transmitted together with the data sub-packet.
Synchronous HARQ transmission entails operating the system on the basis of a predefined sequence of
retransmission packet format and timing.
The benefits of synchronous HARQ operation when compared to asynchronous HARQ operation are:
- Reduction of control signalling overhead. from not signalling the HARQ channel process
number.
- Lower operational complexity if non-adaptive operation is chosen.
Release 7 37 3GPP TR 25.814 V7.1.0 (2006-09)
- Possibility to soft combine control signalling information across retransmissions for enhanced
decoding performance if non-adaptive operation is chosen.
Depending on the actual L1/L2 requirements, asynchronous HARQ may best address the issues of:
- Scheduling flexibility if fully adaptive operation is selected and if both localized and
distributed allocations are selected.
- Support for multiple simultaneous (in the same (set of) sub-frame(s)) independent HARQ
processes
- Flexibility in scheduling of retransmissions
The desirability of particular L1/L2 features will determine the degree of adaptive operation.
7.1.2.4.1 Purposes of the SCH, BCH, and reference symbols and information
to be detected in the cell search
E-UTRA cell search is assumed to be based on two signals (“channels”) transmitted in the downlink,
the “SCH” (Synchronization Channel) and “BCH” (Broadcast Channel).
The primary purpose of the SCH is to enable acquisition of the received timing, i.e., at least the SCH
symbol timing, and frequency of the downlink signal. The UE can obtain the remaining cell/system-
specific information from the BCH, SCH and also from some additional channels, such as the reference
symbols. The primary purpose of the BCH is to broadcast a certain set of cell and/or system-specific
information similar to the current UTRA BCH transport channel.
Aside from the SCH symbol timing and frequency information, the UE must acquire at least the
following cell-specific information.
Each set of information is detected by using one or several of the SCH, reference symbols, or the BCH.
The information regarding the overall transmission bandwidth of the cell can directly indicate the BCH
bandwidth since the BCH bandwidth in relation to the transmission bandwidth of the cell is pre-
specified. To facilitate Cell ID detection, several options in embedding the Cell ID into the SCH are
possible. For example, the Cell ID may be directly mapped into the SCH, or different Cell ID
information may be group-wised. For the case of group ID, cell ID group index can be detected using
the SCH, and the Cell IDs within the detected Cell ID group can be detected using reference symbols
or the BCH. As an alternative approach, information regarding the BCH bandwidth and CP length may
be detected by blind detection from the SCH or BCH, by using hypothesis testing for example.
Detailed information conveyed by the SCH, reference symbols, and BCH should be studied during SI.
Release 7 38 3GPP TR 25.814 V7.1.0 (2006-09)
SCH BCH
τ
Time
10-msec radio frame
SCH BCH
τ
10/n msec
Time
10-msec radio frame
SCH BCH
τ
10/n msec
Time
10-msec radio frame
Figure 7.1.2.4.2-1 – Basic transmission timing of downlink SCH and BCH during one radio
frame. Please note that this figure is for illustrative purposes only and is not meant to specify the
number of SCHs and BCHs per radio frame.
The position of the SCH transmission timing within a sub-frame should be fixed in all sub-frames to
which the SCH is multiplexed and the spacing of the sub-frame with the SCH should be constant, in
order to allow a simple averaging of the correlations over multiple sub-frames. To achieve constant
SCH transmission timing within a sub-frame, the following three SCH symbol multiplexing methods
should be investigated.
frames. The first method allows flexible CP length allocation, i.e., allocation of long and short CPs to
any sub-frame without restrictions from the SCH detection perspective. On the other hand, using this
approach, restrictions on the UL/DL switching point for TDD are necessary in the case that the DTX of
the last part of the downlink sub-frames is used for creation of the TDD guard time.
1 sub-frame
OFDM symbol
CP
Same SCH position regardless of the CP length
Figure 7.1.2.4.2-2 Basic transmission timing of downlink SCH within sub-frame in the first
method. Please note that this figure is for illustrative purposes only.
Figure 7.1.2.4.2-3 shows the second method. In the second method, an SCH symbol is multiplexed into
the first OFDM symbol within a sub-frame and a short CP is used for that OFDM symbol regardless of
the CP length for the other OFDM symbols within that sub-frame. This method also leads to a fixed
SCH symbol transmission timing at the UE and allows flexible CP length allocation and a flexible
UL/DL switching point for the TDD mode. On the other hand, the necessity for SFN reception of the
SCH from multiple cells, and the decoding procedure and transmission performance of the other
channels mapped on the first OFDM symbol with the SCH should be investigated further since the
second method cannot apply a long CP to the SCH when applying a short CP to unicast while applying
a long CP to broadcast for SFN reception.
1 sub-frame
OFDM symbol
Figure 7.1.2.4.2-3 – Basic transmission timing of the downlink SCH within a sub-frame in the
second method. Please note that this figure is for illustrative purposes only.
On the other hand, in the third method, an SCH symbol is multiplexed into any OFDM symbol within
any sub-frame as shown in Fig.7.1.2.4.2-4. In this method, one or more arbitrary OFDM symbols of a
sub-frame may be used for mapping the SCH. However, since the same OFDM symbol position within
a sub-frame and within a radio frame should be maintained, it is necessary to restrict MBMS
transmission to sub-frames not affected by the SCH in the case that CPs of different lengths are used
for the SCH and MBMS. Note that in all the three methods, if SCH transmission occurs multiple times
in a radio frame, the information provided by the SCH may be distributed over multiple transmissions,
or may be repeated.
Release 7 40 3GPP TR 25.814 V7.1.0 (2006-09)
1 sub-frame
OFDM symbol
Figure 7.1.2.4.2-4 – Basic transmission timing of the downlink SCH within a sub-frame in the
third method. Please note that this figure is for illustrative purposes only.
Additionally, the reference symbol detection method when there are sub-frames with different CP
lengths within a radio frame, e.g., for multiplexing the MBMS channel, should be studied. There are
two important issues to be studied.
- Restrict the time multiplexing of long-CP and short-CP sub-frames in order to allow
performance improvements through averaging
- Modify the CP structure of the first OFDM symbol of a long-CP sub-frame in order to keep
the same time-alignment of the first reference signal between short-CP and long-CP sub-
frames (Figure 7.1.2.4.2-5)
1 sub-frame
OFDM symbol
Same reference symbol position Post fix of the first OFDM symbol
regardless of the CP length
Figure 7.1.2.4.2-5 – Transmission timing of the first reference symbol within a sub-frame with
long CP. Please note that this figure is for illustrative purposes only.
- The center frequency of the center sub-carrier over the overall transmission band of each cell
site is to be designed to satisfy the E-UTRA raster condition regardless of the overall
transmission bandwidth of the cell site.
- The downlink SCH is transmitted only in the central part of the overall transmission band of
the cell. As shown in Fig. 7.1.2.4.3-1, working assumption is to focus the study on a SCH
structure based on the constant bandwidth of 1.25 MHz regardless of the overall transmission
bandwidth of the cell, at least for initial cell search.
- The downlink BCH is also transmitted in the central part of the transmission bandwidth of the
cell. In Fig. 7.1.2.4.3-2, the constant bandwidth of 1.25 MHz is used for the downlink BCH,
regardless of the overall transmission bandwidth of the cell. Alternatively, the BCH may be
transmitted over a 5-MHz frequency band to increase the frequency diversity effect when the
overall transmission bandwidth is equal to or wider than 5 MHz. In the latter case, the UE may
Release 7 41 3GPP TR 25.814 V7.1.0 (2006-09)
acquire the BCH bandwidth from the SCH to reduce the decoding complexity of the BCH,
compared to the case when the UE tries to decode all the possible BCH bandwidths. Whether
1.25 MHz or 5 MHz is better as the BCH bandwidth should be investigated during the SI
phase. Additionally, it should be investigated during the SI phase whether parts of the system
information can be broadcast using a 1.25-MHz BCH and then, if necessary, the remainder of
the system information is broadcast using a 5-MHz BCH.
- Regardless of the total transmission bandwidth capability of a Node B, a UE should be able to
determine the cell ID using only the central portion of the bandwidth (i.e., the part of the
bandwidth containing the SCH) in order to achieve very fast cell search.
Note that Figs. 7.1.2.4.3-1 and 7.1.2.4.3-2 are for illustration purposes only and the details of the SCH
and BCH structure should be studied during SI.
20-MHz bandwidth
SCH
10-MHz bandwidth
5-MHz bandwidth
2.5-MHz bandwidth
1.25-MHz bandwidth
Figure 7.1.2.4.3-1 – Frequency allocation of the downlink SCH. Independent of the overall
transmission bandwidth, the SCH is defined for 1.25 MHz and centered in the middle of the
overall transmission bandwidth.
20-MHz bandwidth
BCH
10-MHz bandwidth
5-MHz bandwidth
2.5-MHz bandwidth
1.25-MHz bandwidth
Figure 7.1.2.4.3-2 – Frequency allocation of the downlink BCH. Independent of the overall
transmission bandwidth, the BCH is defined for 1.25 MHz and centered in the middle of the
overall transmission bandwidth.
Release 7 42 3GPP TR 25.814 V7.1.0 (2006-09)
Furthermore, allocation of a larger bandwidth for the SCH and BCH with repetition in the frequency
domain and time-shifting of the block-wised BCH information in the case of a 20-MHz transmission
bandwidth should be considered from the viewpoint of mobility support.
A typical start up procedure would be that a UE first detects the central part of the spectrum regardless
of the receiving bandwidth capability of the UE and the transmission bandwidth of the Node B and
performs a cell search. Then, the UE moves to the transmission bandwidth for actual communications
assigned to it by the system. This procedure is illustrated in Fig. 7.1.2.4.3-3.
Example: 10-MHz UE in 20-MHz cell site, SCH bandwidth = 1.25 MHz and BCH bandwidth = 1.25 MHz
Cell site with 20-MHz transmission bandwidth
BCH
Step 1: SCH
Cell search using synchronization
channel
detect center 1.25 spectrum of
entire 20-MHz spectrum
Step 2: BCH
BCH reception reception
These diversity schemes can also be considered for the BCH in order to improve the packet error rate
(PER) of the BCH. In addition, if the configuration of the transmitter antennas of the cell is provided
outside the BCH, e.g., using the SCH or reference symbols, block code-based transmit diversity can be
considered.
Both SCH signal structures should be studied and compared. The selection of the SCH structure should
be determined taking into account following issues.
- Cell search time performance in presence of inter-cell interference and frequency offset
Release 7 43 3GPP TR 25.814 V7.1.0 (2006-09)
In the hierarchical SCH, two or three signals are used for synchronisation acquisition and determining
the cell ID and possibly other relevant information for the cell or network as mentioned in 7.1.2.4.1. A
primary SCH, using the same OFDM waveform in all cells or a small set of OFDM waveforms with
each cell using one of the OFDM waveforms, is used for SCH symbol timing and frequency
acquisition. A secondary SCH, using a cell specific OFDM waveform, may be used for determining the
cell group ID, full cell ID or other relevant information. If secondary SCH carries just cell group ID,
not full cell ID, the cell specific common reference symbols can be used for determining the full cell
ID. In case no secondary SCH is specified, the reference symbols specify the full cell ID.
In the non-hierarchical SCH, the SCH consists of one or more cell-specific OFDM waveforms. A cell-
specific OFDM waveform can be obtained by IDFT either of a complete cell-specific sequence or
certain portion of a cell-specific sequence, where the sequence elements are used as the Fourier
coefficients at the occupied sub-carrier frequencies. The occupied sub-carrier frequencies may be
different on the different OFDM waveforms. The cell-specific sequence may indicate cell ID or cell
group ID. Each OFDM waveform is preceded by a cyclic prefix. One or more cell-specific OFDM
waveforms are characterized by an exactly or approximately symmetric (centrally symmetric or
periodic) shape of their magnitudes at least in those parts of the SCH intended to be used both for the
synchronization acquisition and the cell ID/cell group ID information transmission.
The hierarchical SCH structures may achieve lower cell search time than the non-hierarchical SCH
structures at low SNRs (e.g. SNR < 0 dB). On the other hand, the performances of non-hierarchical
SCH may become better than the performances of hierarchical SCH at higher SNRs (e.g. SNR > 0 dB).
Start
Frame Cell ID
synchronization identification
End
According to the SCH signal structure described in Section 7.1.2.4.5, there are three different options
for the SCH timing detection methods: cross-correlation based detection (SCH replica-based detection),
auto-correlation based detection, and a hybrid of the two methods.
Cross-correlation based detection is applicable to the hierarchical SCH. The SCH received timing of
the target cell is detected by taking the correlation between the received signal and the cell-common
SCH replica or a small set of distinct SCHs in the time domain.
Auto-correlation based detection is applicable to both the hierarchical SCH and non-hierarchical SCH.
In this case, N exactly or approximately symmetric (centrally symmetric or periodic in their magnitude
shape) waveforms appear within the duration of one or multiple OFDM symbols. Thus, the SCH
symbol timing is detected by taking the auto-correlation of N-1 periodic waveforms of the SCH in the
time domain without information pertaining to the cell-specific SCH waveform.
In the next step of the SCH-symbol timing detection, the radio frame timing must be detected when this
is not directly given by the SCH timing, i.e., if the SCH is transmitted more than once every radio
frame.
There are three different options for radio frame timing detection: SCH based detection, BCH based
detection, and reference signal-based detection.
The SCH based detection is applicable to both the hierarchical SCH and non-hierarchical SCH. With
SCH based detection, the radio frame timing can be estimated by detecting the cell-specific SCH
sequence in the frequency domain employing the SCH symbol timing detected in the previous step.
When the primary SCH and secondary SCH are used in the hierarchical SCH, coherent detection of the
cell-specific secondary SCH can be performed using the primary SCH as a reference signal.
The BCH based detection is applicable to both the hierarchical SCH and non-hierarchical SCH. For
BCH-based frame-timing detection, the frame-timing is detected by decoding the BCH. This may
include hypothesis testing if the BCH is transmitted less frequently than the SCH. This method requires
BCH reception both for the initial cell search and neighboring cell search.
The reference signal based detection is primarily considered for the hierarchical SCH. The frame
timing information is detected by the reference signal waveform (modulation pattern). In this case, the
repetition interval of the reference signal waveform should be equal to the radio frame period, 10 msec.
Following the detection of the SCH-symbol of the target cell, the cell ID must also be detected.
Baseline assumption on the number of cell IDs for evaluation is 512, which is the same as in the current
W-CDMA. However, further study is needed on the number of cell IDs from the viewpoints of flexible
cell-specific sequence assignment and performance. In the cell ID detection step, the following two
Release 7 45 3GPP TR 25.814 V7.1.0 (2006-09)
issues should be decided: what physical channel(s) should be used and the necessity of cell ID
grouping.
Two options are considered for cell ID detection: the SCH or reference symbols. The SCH can be used
for cell ID detection. Here, the SCH sequence directly indicates the cell ID. SCH based cell ID
detection is applicable to both the hierarchical SCH and non-hierarchical SCH. Alternatively, if the
reference signals are modulated with a cell-specific sequence and/or cell-specific hopping pattern that
corresponds to the cell ID, the cell ID can be detected by taking the maximum correlation peak between
the received reference symbols and reference symbol-replica. Reference symbol based cell ID detection
is also applicable to both the hierarchical SCH and non-hierarchical SCH.
The grouping of cell-specific sequences (cell IDs) similar to that in W-CDMA can be applied to cell
search in OFDM based radio access in order to reduce the number of correlation detections. For
example, in the reference symbol based detection of cell IDs, the SCH sequence indicates the cell ID
group. Then, after detecting the cell ID group through SCH sequence detection, only the cell-specific
sequences belonging to the detected group are searched using reference symbols. A similar approach of
grouping (and detecting) cell IDs can be considered for the pure SCH based detection of cell IDs.
Note that other information to be detected in the cell search described in Section 7.1.2.4.1 is detected
after the SCH timing detection step.
- Inter-cell-interference randomization
- Inter-cell-interference cancellation
- Inter-cell-interference co-ordination/avoidance
In addition, the use of beam-forming antenna solutions at the base station is a general method that can
also be seen as a means for downlink inter-cell-interference mitigation.
It should be noted that the different approaches could, at least to some extent, complement each other
i.e. they are not necessarily mutually exclusive.
The possibility to perform inter-cell interference cancellation at the UE is considered irrespective of the
interference mitigation scheme adopted at the transmitter. The radio interface definition should
facilitate the acquisition of channel parameters of a limited number of (strongest) interfering cells (e.g.
through orthogonal reference signals).
A pseudo-random method can be used to generate the cell-specific interleaver patterns for IDMA. The
number of the available patterns (seeds) is determined by the length of interleaver. A UE can identify
the interleaver pattern of the cell by checking its interleaver pattern ID. The seeds can be reused
between “far-spaced” cells in a manner similar to that of frequency reuse in a cellular system.
- Spatial suppression by means of multiple antennas at the UE. It should be noted that the
availability of multiple UE antennas is an assumption for E-UTRA.
- Interference cancellation based on detection/subtraction of the inter-cell interference. One
example is the application of cell-specific interleaving (IDMA) to enable inter-cell-
interference cancellation.
The IDMA based inter-cell-interference cancellation scheme would imply the following requirements
on the system:
1. RB allocation: The RBs accommodating one code block of a UE in the interfered cell should
also accommodate, and only accommodate, a code block of a UE in the interfering cell. In
other words, the “interfered code block” and “interfering code block” should be
accommodated in the same set of RBs.
2. Synchronization: Inter-NodeB synchronization is required.
3. Intra-cell signalling: A UE needs to be signalled whether it can perform a cancellation to the
received ICI. When IDMA is used, the interleaver pattern ID also needs to be signalled to the
UE.
4. Inter-cell signalling: Interfering signal configurations (e.g. interleaver pattern ID, modulation
scheme, FEC scheme and coding rate) should also be signalled to the UE. To cancel the inter-
NodeB interference, the signalling of interfering signal configurations can be realized by
detecting the interfering control channel at the UE. To cancel the inter-sector interference, the
NodeB can straightforwardly signals the interfering signal configurations to the UE via its
own controlling channel.
7.1.2.6.3 Inter-cell-interference co-ordination/avoidance
The common theme of inter-cell-interference co-ordination/avoidance is to apply restrictions to the
downlink resource management (configuration for the common channels and scheduling for the non
common channels) in a coordinated way between cells. These restrictions can be in the form of
restrictions to what time/frequency resources are available to the resource manager or restrictions on
the transmit power that can be applied to certain time/frequency resources. Such restrictions in a cell
will provide the possibility for improvement in SIR, and cell-edge data-rates/coverage, on the
corresponding time/frequency resources in a neighbour cell.
- Alternative #1:
No additional UE measurement and reporting is needed, in addition to CQI reports anyway
needed to support channel-dependent scheduling and link adaptation
- Alternative #2:
Additional UE measurement and reporting of average path loss (incl. shadowing) to current
and neighbour cells. Reporting rate: In the order of once every 100 ms.
- Alternative #3:
In addition to the measurements/reports of alternative #2, additional measurement and
Release 7 47 3GPP TR 25.814 V7.1.0 (2006-09)
reporting of average interference for the frequency reuse sets. Reporting rate: In the order of
once every 100 ms.
7.1.3.1 UE measurements
The UE should be able to measure and report to the Node B the channel quality of one resource block
or a group of resource block, see Section 7.1.3.1.1.1.1 in form of a CQI. In order to allow for efficient
trade-off between UL signaling overhead and link-adaptation/scheduling performance taking varying
channel-conditions and type of scheduling into account, the time granularity of the CQI reporting
should be adjustable in terms of sub-frame units (periodic or triggered) and set on a per UE or per UE-
group basis. In addition, the amount of overhead should be considered when comparing different CQI
reporting schemes. .
The frequency dimension of OFDM symbols can be organized into an integer number of CQI bands
across all carrier bandwidth modes with each CQI band bandwidth corresponding to x (e.g. x=25 or 50)
number of consecutive sub-carriers. The granularity of the CQI band bandwidth should be multiples of
the minimum resource block bandwidth.
Channel quality indicator (CQI) feedback from UE which indicates the downlink channel quality can
be used at Node B at least for the following purposes:
Channel quality measurements defined in section 7.1.3.1.1.1 and some measurements defined in
section 7.1.3.1.2 can be used for interference coordination/management purpose. Additional
measurements for interference coordination/avoidance are considered in sections 7.1.2.6.3 and 9.1.2.7.1
for DL and UL respectively.
1) PLMN selection
a. Detecting available PLMNs and high quality PLMNs
2) Cell selection and cell reselection
a. Detecting suitable cells (i.e., "S-criteria or out-of-range")
b. Determining the most suitable cell for connection, i.e. performing evaluation for cell
reselection
3) Handover decision
a. Intra frequency handover
b. Inter frequency handover
c. Inter RAT handover (GERAN, UTRAN)
d. Measurement gap control (FFS)
Neighbour cell measurements performed by the UE are named intra-frequency measurements when the
UE can carry out the measurements without re-tuning its receiver. This corresponds to the case when
the current and target cell operates on the same carrier frequency and
- the UE maximum bandwidth capability is equal or larger than the network system bandwidth,
or
- the UE maximum bandwidth capability is smaller than the network system bandwidth, but the
UE is currently “camping” within a band so that the common channels of the target cell as
well as the sub-carriers allocated to the UE on the DL shared channel are within the UE
receiving bandwidth.
Note that the exact definition of the common channels that are needed for the neighbour measurements
as well as the meaning of “camping” are not yet fully agreed upon.
Neighbour cell measurements are considered inter-frequency measurements when the UE needs to re-
tune its receiver in order to carry out the measurements. This corresponds to the cases when
- the neighbour cell is operating on a different carrier frequency than the current cell, or
- the neighbour cell is operating on the same carrier frequency as the current cell, but the UE
maximum bandwidth capability is smaller than the network system bandwidth and the UE is
currently “camping” within a band so that the common channels of the target cell are outside
Release 7 49 3GPP TR 25.814 V7.1.0 (2006-09)
the UE receiving bandwidth. Note that depending on the exact structure for the common
channels needed for neighbour measurements, this scenario may not happen.
In case of inter-frequency measurements, the network needs to be able to provide UL/DL idle periods
for the UE to perform necessary neighbour measurements.
Neighbour measurements are considered inter-RAT measurements when UE needs to measure other
radio access technology cells. For these kinds of measurements, the network needs to be able to provide
UL/DL idle periods.
In case the UE needs UL/DL idle periods for making neighbour measurements or inter-RAT
measurements, the network needs provide enough idle periods for the UE to perform the requested
measurements. Such idle periods are created by the scheduler, i.e. compressed mode is assumed not
needed. In order to optimise the network, some additional measurements may be used by the network
for triggering the generation of UL/DL periods. This is FFS.
• Frequency re-use 1
• Node-B scheduling
The following additions to the baseline multiple access structure should be considered:
• Support for simultaneous reception of HS-DSCH and multicast data transmitted according to the
simulcast structure and procedures described in clause [X] should be evaluated.
• Added support for 0.96 Mcps and possibly 1.92 Mcps numerology.
• Enhanced downlink control structure and procedures in support of HS-DSCH and E-DCH
operation with variable symmetric and asymmetric bandwidth allocations.
The system operation should rely on the definition of new demodulation performance requirements as
follows:
• Enhanced reference equalizer such as frequency domain equalizer with decision feedback
combined with multi-antenna receive diversity in the UE.
The concepts presented are valid for multi-carrier operation based on the 5MHz system (UTRA FDD)
as well as the 1.25MHz system (WCDMA LCR) or a 5MHz/1.25MHz hybrid multi-carrier system.
7.2.2.1 Definitions
N: maximum number of DL carriers that a UE may receive at a given time. The notion of “carrier
assignment” is implicit with this concept, i.e., the E-UTRAN shall notify the UE that it may receive
data on up to N carriers simultaneously.
M: maximum number of UL carriers that a UE may transmit at a given time. Also, the notion of
“carrier assignment” is implicit with this concept, i.e., the E-UTRAN shall notify the UE that it may
transmit data on up to M carriers simultaneously.
Paired Carriers: carrier frequencies such that for each DL carrier there is an associated UL carrier.
The PHY channel timing relationships for paired carriers are the same as those for the currently defined
single carrier UMTS system.
Unpaired Carriers: carrier frequencies that do not have an associated DL carrier (in case of M>N) or
do not have an associated UL carrier (in case of N>M). The timing of PHY channels of unpaired
carriers will be associated with the timing of some paired carriers. Details on this will be provided in
the body of this document.
Anchor Carrier: carrier within a cell that contains full R99, R5 or R6 capability, i.e., transmission of
SCH, P-CCPCH, S-CCPCH… and supporting reception of UE random access by way of the PRACH.
For the downlink, the anchor carrier in a cell constitutes the timing reference for all the carriers in that
cell i.e., time synchronization within the carriers in a cell is assumed as stated section 7.2.2.2.
Parameter Value
Carrier spacing 1.25MHz
Chip rate 960 kcps
Slot duration 0.67ms: 640 chips
Max SF 128
R5/R6 TTIs 2ms: 3 slots
10ms: 15 slots
The chip rate for the WCDMA LCR system comes from a direct translation to a 1.25MHz carrier
spacing of the relationship between the carrier spacing and the chip rate for the UTRA FDD, i.e.,
3.84Mcps is to 5MHz the same as 960kcps is to 1.25MHz. Therefore, the same pulse shaping filter taps
are assumed for the 1.25MHz system, with the difference that the sample duration is 4 times longer in
the 1.25MHz system.
The slot duration for the 1.25MHz system corresponds to the UTRA slot duration of 0.67ms.
Therefore, since the “chips” of the 1.25MHz system are 4 times longer, the number of chips in a slot is
2560/4=640. The number 640=128x5, and therefore the maximum spreading factor is 128, for which
there will be 5 Walsh symbols per slot. For this slot duration, the TTI duration in slots corresponding to
the R5 and R6 channels remains the same i.e., 2ms and 10ms, as in UTRA FDD.
The time reference for cells connected to the same Node B is assumed to be common across all
carriers. Therefore, the DL timing reference i.e., timing of the P-CCPCH or SCH is the same for all
carriers in a given cell.
The timing of the PHY channels for paired carriers shall be no different than for a single carrier
system where the timing of all the DL channels is referenced to the timing of the P-CCPH or SCH.
The timing of the PHY channels for unpaired carriers is explicitly covered in this Technical Report.
One cell is the serving HS-DSCH for all carriers supported by a given UE.
HARQ PHY re-transmissions on DL takes place at the same carrier as for the first
transmission.
Figure 7.2.2.3-1 is a block diagram depicting multi-carrier operation. Each of the colours represents a
different DL carrier. Transmission by the serving cell is represented by solid lines, whereas
transmissions by other cells are represented by dashed lines. Note that the PHY channels in squared
brackets are just transmitted if associated uplink carrier is configured.
Release 7 52 3GPP TR 25.814 V7.1.0 (2006-09)
Node-B 1
* HS-SCCH [*F-DPCH]
Controller * HS-DSCH [*E-HICH/E-RGCH]
[* E-HICH/E-RGCH]
[* F-DPCH]
[* E-AGCH]
Node-B 3
UE
Node-B 2
P-CCPCH: Primary Common Control Physical channel carrying the system information i.e.
BCH: Broadcast transport channel.
S-CCPCH: Secondary Common Control Physical channel carrying the Paging (PCH) and the
Forward Access Channel (FACH) transport channels. If there is interest in increased data
transmission capabilities over FACH, transmission of the S-CCPCH could take place on
more than one carrier in a cell.
o PICH: Paging Indicator Channel if S-CCPCH carrying the PCH is transmitted over a
single carrier.
HS-PDSCH N
The Absolute Grant messages for a multi-carrier UE with M uplink carriers may be transmitted on M
independent E-AGCH PHY channels (in same or different carriers) or could be transmitted on a single
PHY channel at a particular DL carrier. In that case, the E-RNTI identifier would require the additional
notion of “carrier” on top of the notion of UE identity. Therefore, a UE could have more than one
associated E-RNTI e.g., one for each UL carrier that is allowed to use.
N=M: All the DL carriers have an associated UL carrier and vice-versa. PHY procedures for
this case (i.e., Power Control, synchronization, HS-DSCH and E-DCH related procedures…)
are no different than those for the single carrier case.
N>M: Just the M paired carriers require (F-)DPCH, E-HICH/E-RGCH and E-AGCH (in case
of M AGCH channels used). There are (N-M) downlink unpaired carriers carrying the HS-
PDSCH and its associated HS-SCCHs.
N<M: There are (M-N) uplink unpaired carriers. Therefore, the following will be required:
o (M-N) additional (F-)DPCHs for UL power control of the (M-N) uplink unpaired
carriers.
Release 7 54 3GPP TR 25.814 V7.1.0 (2006-09)
o (M-N) additional E-AGCH channels for transmission of the absolute grants of the
(M-N) uplink unpaired carriers if M E-AGCHs are transmitted for a UE with M
uplink carriers. If a single E-AGCH is used for the absolute grants of a UE in all the
UL carriers, no additional E-AGCHs are needed.
The (M-N) sets of additional channels ((F-)DPCH, E-HICH/E-RGCH and optionally
E-AGCH) are related to E-DCH transmissions on the UL. Therefore, the cells in the UE’s
E-DCH Active Set of each carrier shall transmit to the UE the supporting E-DCH
feedback information and the RL TPC commands. For cells belonging to the same Node-
B, the transmission of these channels shall take place at the same carriers. It may also be
desired, for implementation reasons, that the carriers for transmission of these channels
are the same for different Node-Bs.
Note that multiple F-DPCHs on a given carrier may be orthogonally time multiplexed within the same
channelisation code by using different timing offsets (multiples of 256 chips). Therefore, the additional
F-DPCHs may be time multiplexed within a set of DL carriers. Alternatively, different channelisation
codes may be used for the additional F-DPCHs with timing same or different than that of the paired F-
DPCH.
This section covers the timing specifics of DL PHY channels for asymmetric, i.e., N ≠ M multi-
carrier configurations.
In the N>M case, there are (N-M) downlink unpaired carriers. The timing of the DL channels in these
carriers (HS-PDSCH and HS-SCCH) is well defined since for the DL, the timing of all the PHY
changes is referenced to the nominal timing of the P-CCPCH or SCH of the anchor carrier.
In the N<M case, there are (M-N) uplink unpaired carriers. Also, there are (M-N) sets of channels ((F-
)DPCH, E-HICH/E-RGCH and possibly E-AGCH) that need to be allocated within the N downlink
carriers. The timing of E-HICH is indirectly related to the timing of the associated (F-)DPCH. The
timing of the E-RGCH for the serving cell coincides with the timing of the E-HICH. Whereas, the
timing of the E-RGCH from a non-serving cell as well as the timing of the E-AGCH is absolute with
respect to the nominal timing (2 slots after). In addition, as noted before, the E-AGCH may be
transmitted on a single carrier. Therefore, the (M-N) additional (F-)DPCHs (on top of the N ones
corresponding to the paired carriers) will have a particular timing multiple of 256 chips which will
constitute the indirect reference for the E-HICH and the E-RGCH from the serving cell.
Note that multiple F-DPCHs on a given carrier may be orthogonally time multiplexed within the same
channelization code by using different timing offsets (multiples of 256 chips). Therefore, the additional
F-DPCHs may be time multiplexed within a set of DL carriers. Alternatively, different channelization
codes may be used for the additional F-DPCHs with timing same or different than that of the paired F-
DPCH.
Release 7 55 3GPP TR 25.814 V7.1.0 (2006-09)
There are no changes in the procedure described in 25.214, however just a subset of carriers (the
smallest subset being the single anchor carrier) will carry the P-SCH/S-SCH and the P-CCPCH and
therefore will enable the UE to perform the three steps for system acquisition.
If different carriers from the same cell share a common timing reference (as assumed in section
7.2.2.2), there is no need for steps 1 and 2 in the system acquisition process described in 25.214
(related to the acquisition of the slot and frame timing as well as the identification of the scrambling
code group to which the cell belongs, through acquisition of P-SCH and identification of S-SCH). Step
3 (last step) in the system acquisition procedure described in 25.214 may also be avoided if different
carriers from different cells use the same scrambling code.
Common timing reference for all carriers from same cell, and
If the common timing reference for carriers from same cell was optional. A signalling message could
be defined with the objective of indicating whether or not each of the DL carriers share the same timing
as the anchor carrier.
In turn, if the use of common scrambling code for carriers from same cell was optional. The E-UTRAN
would have to indicate through signalling (e.g., P-CCPCH or S-CCPCH) the scrambling code used by
carriers other than the anchor carrier. This approach would add unnecessary overhead on the DL and is
not desired.
7.2.3.3.1 Procedure A
This section covers Procedure A step by step identifying the points that could be different for the
addition of a new carrier.
Procedure A step “b” in 25.214 currently specifies that the initial transmit power for the DL DPCCH or
F-DPCH is set by higher layers. This transmit power could be chosen to be the same as that for one of
the established carriers.
The DL chip and frame synchronization described in step “c” may be simplified for the assumption of
common timing from different carriers.
Step “d” specifies when the UE may start transmitting. Currently, higher layers need to consider the DL
physical channel established and activation time (if provided) reached. After that, transmission of
DPCCH start at an initial transmit power set by higher layers. This initial DPCCH power may also be
the same as one of the other active carrier’s DPCCH. The power control preamble, in this case, may
also be reduced to speed up the synchronization.
Release 7 56 3GPP TR 25.814 V7.1.0 (2006-09)
7.2.3.3.2 Procedure B
Synchronization Procedure B does not involve the UE as it governs the addition of radio links to the
existing radio link sets between the UE and E-UTRAN. Therefore, it would not require any change to
specifically support multi-carrier operation.
M=N: each UL carrier has its associated DL carrier and vice-versa. Therefore, the M uplink
DPCCHs will power control the M downlink (F-)DPCHs, and optionally the HS-SCCHs and
the DL E-channels.
M>N: there are N uplink DPCCHs meaningful for DL power control. Those are the N paired
carriers that power control the downlink (F-)DPCHs and optionally HS-SCCHs in the N
downlink carriers.
M<N: the DPCCHs in the M paired carriers will power control the M downlink (F-)DPCHs.
Power control of the HS-SCCHs in the (N-M) downlink unpaired carriers may be based on
CQI reports for each of the carriers by the UE.
7.2.3.5 HS-DSCH Related Procedures
Transmission of HSDPA related channels is covered in section 7.2.2.5 for the DL (i.e., HS-PDSCH and
HS-SCCH) and section 9.3.2.4 for the UL (i.e., HS-DPCCH).
N=M: each DL carrier has its associated UL carrier and vice-versa. Therefore, the N downlink
HS-PDSCHs/HS-SCCHs will be fed back by the corresponding N uplink HS-DPCCHs.
N>M: the HS-PDSCHs/HS-SCCHs in the M paired carriers will be fed back by the
corresponding M uplink paired HS-DPCHs. In addition, the UE further conveys CQI and
ACK/NACK commands for the (N-M) downlink unpaired carriers for HARQ operation and
channel feedback of the HS-PDSCH of those carriers. How that additional information is
conveyed is covered in section 9.3.2.4.1.
N<M: the HS-PDSCHs in the N paired carriers will be fed back by their corresponding HS-
DPCCHs in the paired carriers.
7.2.3.5.1 HS-PDSCH Retransmission on Multi-Carrier system
Operation in the multi-carrier system shall guarantee PHY HARQ retransmissions on the carrier that
was used for the first transmission.
7.2.4.1 UE measurements
The UE measurements for the MC-WCDMA based proposal are the same as those defined in section
5.1 of 25.215.
Release 7 57 3GPP TR 25.814 V7.1.0 (2006-09)
Likewise, measurements and procedures to support mobility towards and within E-UTRA do not differ
from currently specified procedures.
Measurements periods to support inter-frequency and inter-RAT handover are provided by compressed
mode operation and/or new DTX/DRX procedures for the UE.
For the DL multiple access, the bandwidth of each downlink sub-carrier will be allocated as 1.6 MHz.
s ubfra me (5 ms )
Figure 7.3.1-1: Physical channel signal format for 1.28Mcps TDD option
Switching Point
1.28Mcps
The sub-carrier number of DL and UL in the different bandwidths are given in Table 7.3.1-1.
Table 7.3.1-1 the number of UL and DL sub-carriers according to different bandwidths
MC TD-SCDMA should have similar structure as LCR TDD (Figure 7.3.1-1), i.e. three layered
structure: radio frame, sub-frame and time slot. The number and position of traffic and special time
slots are also the same, aligned with current system. The TTI is 0.675ms, the same size as the period of
a timeslot.
For general services, MC TD-SCDMA can use the same number of switching points as shown in
Figure 7.3.1-2.
More Switching points can be added in the 5ms sub-frame of MC TD-SCDMA to meet special
requirements as shortened latency. For example, one pair of switching points can be added as in Figure
7.3.1-4 to meet the 5ms unidirectional user plane latency requirement in the LTE. In Figure 7.3.1-4,
TS0, TS2, TS3, TS5 and TS6 are downlink timeslots, while TS1, TS4 act as uplink. One pair of
switching points is added to reduce latency.
In Figure 7.3.1-4, TS4 is composed of two parts: the guard period GP and T4. The functionality of GP
is the same as special time slot GP, which acts as a protection between downlink and uplink slots. T4 is
a part which used to transmit uplink data. To maintain the alignment with LCR TDD system, the length
of TS4 remains 0.675ms. The number of switching points can be further increased on the basis of
Figure 7.3.1-4 to fulfil even more stringent needs for lower latency.
DUSP
UDSP UDSP
DwPTS UpPTS
GP GP T4
DUSP
There are two main multiple antenna technologies, namely Beam-forming and MIMO. MC TD-
SCDMA combines these two technologies according to different application environment and channel
characteristics.
The baseline antenna configuration for downlink MIMO is to deploy two transmit antennas at the Cell
site, and two receive antennas at the UE. The possibility for more transmit antennas should also be
considered.
The antenna configuration for downlink beam-forming is to use from four to eight transmit antennas at
the Cell site, and one or two receive antennas at the UE.
7.3.2.1 Scheduling
For DL MC TD-SCDMA, frequency diversity, time diversity and space diversity should be considered.
Using AMC to adjust the modulation and coding rate, adaptive link technologies improve the
performance of the system.
7.3.2.3 HARQ
Incremental Redundancy (IR) should be used for downlink HARQ. Note that Chase combining is a
special case of IR.
The open-loop and close-loop power control are supported against deep fading, eliminating near-far
effect, and fighting multi-user interference.
Release 7 60 3GPP TR 25.814 V7.1.0 (2006-09)
7.3.3.1 UE measurements
Table 8.1.1-1 – Traffic Outage and Latency requirements for determining maximum load
Traffic Type Outage Limit and Definition
HTTP – Web Browsing with TCP 2% outage based on user packet call throughput < P
FTP – with TCP 2% outage based on user packet call throughput < Q
VoIP 2% outage based on user having < 98% of its speech frames
delivered successfully within [40] ms (air interface delay).
S Kbps Streaming Video 2% outage based on user having > 2% dropped packets
This section provides initial results of different E-UTRA downlink proposals, comparing with the
baseline reference case defined in [4]:
- WCDMA Release-6
- Rake receiver
Figure 8.1.2.2.1-1 presents achievable throughput as a function of the received Es/N0 assuming a 2x2
MIMO configuration for two types of UE receiver and various modulation and code rates.
120
QRM-MLD using ASESS
MMSE
100
QPSK R = 1/2
QPSK R = 2/3
QPSK R = 3/4
Throughput (Mbps)
80 16QAM R = 1/2
16QAM R = 2/3
16QAM R = 3/4
16QAM R = 4/5
60 64QAM R = 2/3
64QAM R = 3/4
64QAM R = 4/5
40
20
0
0 5 10 15 20 25 30
Average received Es/N0 per receiver branch (dB)
*
System B System C
System A
*
* *
* *
(1,2) (2,2) (2,4) (4,4) (1,2) (2,2) (2,4) (4,4) (1,2) (2,2) (2,4) (4,4) Formatted: Font: 9 pt
Figure 8.1.2.2.1-2: peak rate, mean rate, and cell edge rate performance metrics (iid)
Figure 8.1.2.2.1-3 shows the user throughput distribution for a scenario with 6 dB additional isolation
between cells using 2x2 MIMO with and without pre-coding.
0.9
0.8
Fraction of Positions with Throughput < X
0.7
0.6
0.5
0.4
0.3
0.2
Figure 8.1.2.2.1-3: 10 MHz, 500 m ISD, 20 dB penetration loss, SCM-E Urban, 3 km/h
Results presented in figures 81.12.2.1-1,2 and 3 show that the downlink parameters assumed for the
evaluation are sensible given the achievable peak user rates in various deployment scenarios.
25.913 sections 7.1 and 7.2 and is based on the E-UTRA (OFDMA) and WCDMA reference Node-B
and UE configurations described in TR 25.913 and Annex A.
The initial evaluation results presented in Tables 8.1.2.2.1-1, 8.1.2.2.1-2 and 8.1.2.2.1-3 are based on
full buffer traffic models and proportional fair scheduler. It is assumed that the scheduler is able to
independently allocate individual sub-bands to different UEs at the same time. Further, it is assumed
that the UE reports full CQI for all downlink sub-bands.
The downlink overhead for OFDM results in Tables 8.1.2.2.1-1, 8.1.2.2.1-2 and 8.1.2.2.1-3 is assumed
to be 25%, 29% and 20% respectively.
A summary of the downlink reference system evaluation baseline results of cell and user throughput
performance relative to the TR 25.913 WCDMA reference (e.g. Type I HSDPA UE) for deployment
cases 1, 2, and 3 are given in Table 8.1.2.2.2.2-2 and 8.1.2.2.2.2-3. Table 8.1.2.2.2.2-1 indicates some
key simulation assumptions used by each source. Note, TR 25.913 sections 7.1 and 7.2 give the
relative cell and user throughput target requirements and the LTE and WCDMA reference UE and
Node-B configurations. More details on reference UE and Node-B assumptions can be found in
Annex A.
Key Simulation (1) (2a) (2b) (3a) (3b) (4a) (4b) (4c) (5) (6) (7)
Assumptions R1-062399 R1-061626 R1-061626 R1-061238 R1-062358 R1-061381 R1-061865 R1-061865 R1-061581 R1-061281 R1-061507,28
Overhead 25% 29% 19% 29% 29% 29% 29% 11% 29% 29% 29%
Channel EstImation Non-ideal Ideal Ideal Non-ideal Non-ideal Ideal Ideal Ideal Ideal Non-Ideal Ideal
2x2 MIMO Yes Yes Yes No Yes PARC EBF EBF PARC No No
Receiver Type PU2RC QML QML 2RX-IRC LMMSE IRC-SIC IRC-SIC IRC-SIC MMSE-SIC MRC LMMSE
TTI 0.5ms 0.5ms 1.5ms 0.5ms 0.5ms 0.5ms 0.5ms 2ms 0.5ms 0.5ms 0.5ms
CQI Delay 1.5ms 1.5ms 1.5ms 4.0ms 4.0ms 0ms 0ms 0ms 1.0ms 2.0 ms 1.5ms
CQI Reporting Interval 4.0ms 2.0ms 1.5ms 0.5ms 0.5ms 0.5ms 0.5ms 0.5ms 0.5ms 0.5ms 1.5ms
TTI MUXing up to 8 up to 8 1UE/RB 1UE/RB 1UE/TTI 1UE/TTI 1UE/TTI up to 20 up to 24
#UEs/Sector 32 10 10 48 48 1 1 1 20 10 10
Reuse 1x3x1 1x3x1 1x3x1 1x3x1 1x3x1 1x3x1 1x3x1 1x3x1 1x3x1 1x3x1 1x3x1
Traffic Type FB FB FB FB FB On/Off On/Off On/Off FB On/Off FB
FDS Used Yes/375kHz Yes/375kHz Yes/375kHz Yes/375kHz Yes/375kHz Yes Yes Yes Yes/375kHz Yes/375kHz Yes/750kHz
Fairness Info Yes Yes Yes Yes Yes Yes Yes Yes
I.M.,
Possible Further CLMIMO, CLMIMO, 2ms TTI, I.M.
IC, IRC FFR,SFR FFR,SFR 2ms TTI 4x2 MIMO MIMO, FDS
Improvements VarTTI VarTTI 4x2 MIMO R1-061444
VarTTI
Case 1: Sector 0,748 0,812 0,628 0,44 0,97 0,59 0,988 2,16 2,36 2,60 1,65 1,88 1,32 1,40 1,60 2,72 1,33 1,68
Case 2: Sector - 0,801 0,446 - - 0,50 0,664 - 1,85 2,13 1,01 - - - - - 0,99 1,40
Case 3: Sector 0,652 0,795 0,599 0,41 0,532 0,55 - 2,22 2,38 1,56 1,77 1,22 1,30 1,60 1,41 1,23 -
Case 1: AvgUser - 0,081 0,026 0,42 0,079 0,27 0,100 - 0,240 0,263 0,069 0,078 1,26 1,4 1,9 0,22 0,99 0,168
Case 2: AvgUser - 0,080 0,019 - - 0,23 0,066 - 0,188 0,216 0,042 - - - - - 0,90 0,140
Case 3: AvgUser - 0,080 0,025 0,40 0,038 0,23 - 0,225 0,241 0,065 0,074 1,18 1,20 1,54 0,10 0,84 -
Case 1: 5%User 0,006 0,022 0,010 0,14 0,020 0,13 0,040 0,018 0,048 0,053 0,021 0,020 0,27 0,37 0,50 0,06 0,34 0,059
Case 2: 5%User - 0,025 0,008 - - 0,11 0,027 - 0,035 0,047 0,010 - - - - - 0,32 0,049
Case 3: 5%User 0,001 0,018 0,007 0,13 - 0,09 - 0,043 0,047 0,017 0,015 0,17 0,27 0,36 - 0,20 -
Release 7 66 3GPP TR 25.814 V7.1.0 (2006-09)
Downlink reference system evaluation baseline performance gain results given in Table 8.1.2.2.2-3
show the E-UTRA OFDMA based downlink relative improvement to the WCDMA 25.913 reference
exceeding 3x for sector and average user throughput and 2x for 5%-ile user throughput for some of the
sources. Simultaneously achieving 3-4x sector and average user throughput and 2-3x 5%-ile user
throughput is difficult but is achieved by sources (4a-c) and (2b) given the use of 2x2 MIMO and (in
the case of 2b and 4c) longer TTI with less control overhead for cases 1 and 3. Further gain
improvement for simultaneously achieving the targets is expected from other features (see last row of
Table 8.1.2.2.2-1).
This section summarizes the amount of relative performance improvement by using 4 transmit antennas
and 2 or 4 receive antennas with respect to the reference case of 2 transmit and 2 receive antennas.
Table 8.1.2.2.2.2.1-1 and 8.1.2.2.2.2.1-2 present the results for 4x2 and 4x4 antenna configurations,
respectively.
Table 8.1.2.2.2.2.1-1. Summary of gains by 4x2 MIMO (4Tx, 2Rx) w.r.t. 2x2 MIMO (2Tx,
2Rx)
Simulation Cases
T-put Metric (1) R1-061865 (2) R1-062045 (3) R1-062358 (4) R1-062399
Type
Case 2: Sector - 6% - -
Case 2: 5% User - 7% - -
(3) 4x2 Unitary Precoding. With RR scheduler, sector T-put gains are 23% (case 1) and 19% (case
3), and 5%-user T-put gains are 53% (case 1) and 55% (case 3).
(4) 4x2 PU2RC. With 2 Tx and 4 Rx, sector T-put gain is 31% and 5%-user T-put gain is 44%.
Table 8.1.2.2.2.2.1-2. Summary of gains of by 4x4 MIMO (4Tx, 4Rx) w.r.t. 2x2 MIMO (2Tx,
2Rx)
Case 1: Sector 59 % 83 % 76 %
Case 2: Sector - 86 % -
Case 3: Sector 56 % 85 % -
(1) 4x4 PGRC. Frequency selective scheduling. With FFR, sector T-put gains are 47% (case 1) and
43% (case 3), and 5%-user T-put gains are 115% (case 1) and 145% (case 3).
Table 8.1.2.2.2.3-1 below summarises the additional enhancement techniques for the downlink with
indicative additional gains to the results in table 8.2.2.2.2-3.
35% 0% R1-061455/R1-062365
35%/0% 0%/65% CONSENSUS
Interference 13% 32% R1-061718 – w/o int. coord
cancellation 19%/33% 33%/64% R1-061894 – w/o int.
(Perfect removal) coord; 1,2 cells cancelled
- 21% R-061444 – w/o int. Coord
>100% >200% R1-062301 - with coord,
tilting and intra-NB
cooperation
15% 32% CONSENSUS – w/o int.
Coord
Adaptive multiple 15%-18% 15%-18% R1-061865
antenna schemes (for 10-18% 0% R1-062145
(2,2) case) 15% 10% CONSENSUS
Semi-static inference 0% 10% (uniform load R1-060368
co-ordination and cell size) -50%
0% (non-uniform)
10% R1-060983
0% 10% CONSENSUS (uniform)
The gains in the sector spectral efficiency and 5%-tile user throughput columns are achieved
simultaneously. It must be noted that 5%-tile user throughput and sector throughput can always be
traded off by the scheduler. Without performing simulations containing all these techniques it is
somewhat difficult to conclude on an exact cumulative gain figures for all these techniques. However it
is likely that with these techniques included in LTE specification and implementation that at least 3.5x
gain in sector spectral efficiency together with at least 3x gain in 5%-tile user throughput can be
achieved. There is the potential, considering some of the simulation results, that even higher figures
could be achieved. However noting that some ideal assumptions are i
(1) Round Robin scheduler amongst users with data in their buffer..
Table 8.1.2.2.3-2 Downlink VoIP capacity for E-UTRA (AMR 12.2 Kbps, 5% outage)
Table 8.1.2.2.3-3 Downlink VoIP capacity for E-UTRA (AMR 7.95 Kbps, 5% outage)
OFDM SFN
Site to
Multicast
Band site Speed
Case 1% BLER
(MHz) distance (kph)
95% coverage
(m)
(Mbps)
> 5.5
1 2000 500 3
(1.1 b/s/Hz)
> 5.5
2 2000 500 30
(1.1 b/s/Hz)
< 2.5
3 2000 1732 3
(0.5 b/s/Hz)
> 5.5
4 900 1000 3
(1.1 b/s/Hz)
OFDMA based multicast offers potential for significant gains over Rel-6 MBMS. However, these
performance gains rely on tightly time synchronized transmission from all cells and are highly sensitive
to the combination of band of operation and the site to site distance as illustrated by the performance
for case 3 as compared to other cases.
8.2.1.1 Baseband
The baseband complexity can be divided in terms of memory and processing. The processing is
dominated by the decoder and the receiver front end (user separation, demodulation and channel
equalization). The complexity of the decoder is related to the peak data rate. For CDMA based signal
the complexity of the receiver front end is essentially linked to the channel equalization.
Receiver front end complexity evaluation for receiver based on frequency domain equalizer shows that
the UE has to perform two FFT operations; furthermore a decision feedback can be added to further
enhance the equalization performance; a first order measure of the decision feedback block complexity
indicates that the complexity is about the same as the complexity of an FFT operation.
One should note that the performance enhancements associated with advance equalizer receivers will
also benefit systems which operate Release 5 UTRA-FDD (HS-PDSCH).
8.3.1.1 Baseband
The WCDMA based E-UTRA downlink relies on the existing WCDMA channel structure and
procedures and therefore Release 6 Node B should be compatible with the E-UTRA channels structure.
The multi-carrier component affects mostly the scheduler and interfaces between the channel elements
and the scheduler. The scheduler has to simultaneously control resource allocation across multiple
carriers for a UE instead of one per UE in Release 6. The interface from and to each channel element
with the scheduler may or may not have to be modified depending on the existing implementation and
whether joint scheduling across carriers is supported. No fundamental Node-B complexity issue has
been identified for the WCDMA based E-UTRA downlink.
Release 7 71 3GPP TR 25.814 V7.1.0 (2006-09)
9 UL Concepts
Four basic concepts are proposed in uplink:
3 MC-WCDMA(FDD)
4 MC-TD-SCDMA (TDD)
Sub-carrier CP
DFT Mapping IFFT insertion
NTX symbols
Size-NTX Size-NFFT
The sub-carrier mapping determines which part of the spectrum that is used for transmission by
inserting a suitable number of zeros at the upper and/or lower end in Figure 9.1.1-2. Between each DFT
output sample L-1 zeros are inserted. A mapping with L=1 corresponds to localized transmissions, i.e.,
transmissions where the DFT outputs are mapped to consecutive sub-carriers. With L>1, distributed
transmissions result, which are considered as a complement to localized transmissions for additional
frequency diversity.
Release 7 72 3GPP TR 25.814 V7.1.0 (2006-09)
0
0
L-1 zeros
L-1 zeros
0
BW
fc
f
1 Nn N
The basic sub-frame structure for the uplink transmission is given in Figure 9.1.1-4 using two short
blocks (SB) and six long blocks (LB) per sub-frame. Short blocks are used for reference signals for
coherent demodulation and/or control/data transmission. Long blocks are used for control and/or data
transmission. Note that the data could include either or both of scheduled data transmission and
contention based data transmission. Furthermore, the same sub-frame structure is used for both
localized and distributed transmission.
The numerology for the different spectrum allocations is shown in Table 9.1.1-1. The minimum TTI for
uplink transmission is equal to the uplink sub-frame duration. Similar to the downlink, the possibility to
concatenate multiple sub-frames into longer uplink TTIs should be considered. In this case, the TTI can
either be a semi-static or dynamic transport channel attribute. In case of a semi-static TTI, the TTI is set
through higher layer signalling. In case of a dynamic TTI, the number of sub-frames concatenated can
be dynamically varied for at least the initial transmission and possibly for retransmissions. It is to be
determined to what extent a dynamic TTI can reduce higher layer protocol overhead (e.g. MAC, RLC),
L1 overhead (e.g. CRC), and ACK/NACK feedback, as well as reducing latency by reducing
segmentation of IP packets. It is initially assumed that the Network (e.g. Node-B) would control the
TTI. The interaction between dynamic TTI, signaling errors, HARQ procedure (time synchronous vs.
asynchronous including adaptive or non-adaptive characteristics) and UE complexity needs to be
investigated.
Release 7 73 3GPP TR 25.814 V7.1.0 (2006-09)
.-
1 sub-frame = 0.5 msec
SB SB
CP LB#1 CP CP LB #2 CP LB #3 CP LB #4 CP LB #5 CP CP LB#6
#1 #2
Table 9.1.1-1. Parameters for Uplink Transmission Scheme using BW efficiency of ~90%
*1: {(x1/y1) × n1, (x2/y2) × n2} means (x1/y1) for n1 reference signal or data blocks and (x2/y2) for n2
reference signal or data blocks
Note that the largest CP duration includes guard time for ramp up + ramp down time
For E-UTRA TDD, the frame structure corresponding to Table 9.1.1-1 is supported. In addition, a
second frame structure is also supported with the intention of providing co-existence with LCR UTRA
TDD. The sampling frequency, FFT size, sub-carrier spacing, and number of occupied sub-carriers is
the same as for Table 9.1.1-1. However, with this alternative frame structure described in section
6.2.1.1, the basic timeslot structure for the uplink transmission is given in Figure 9.1.1-5 using two
short blocks (SB) and eight long blocks (LB) per timeslot. The uplink numerology for the different
spectrum allocations are listed in Table 9.1.1-2.
.-
Table 9.1.1-2 Parameters for Uplink Transmission Scheme using BW efficiency of ~90%
Spectrum Timeslot Long block size Short block size CP duration Timeslot Interval
Allocation duration (µs/#of occupied (µs/#of occupied (µs/samples *1)
(us/samples)
(MHz) (ms) subcarriers subcarriers /samples)
/samples*2)
20 0.675 66.67/1200/2048 33.33/600/1024 (6.71/206) × 9, 7.68/236
(6.97/214) × 1*
15 0.675 66.67/900/1536 33.33/450/768 (6.77/156) × 9, 6.94/160
(7.11/164) × 1*
10 0.675 66.67/600/1024 33.33/300/512 (6.71/103) × 9, 7.42/114
(7.22/111) × 1*
5 0.675 66.67/300/512 33.33/150/256 (6.64/51) × 9, 7.54/58
(7.67/59) × 1*
2.5 0.675 66.67/150/256 33.33/75/128 (6.51/25) × 9, 7.80/30
(8.58/33) × 1*
1.25 0.675 66.67/75/128 33.33/38/64 (6.25/12) × 9, 8.32/16
(10.4/20) × 1*1
1
* : {(x1/y1) × n1, (x2/y2) × n2} means (x1/y1) for n1 reference signal or data blocks and (x2/y2) for n2
reference signal or data blocks
*2: FFT size = samples
Table9.1.1.2.1-1 Bandwidth occupied by a resource unit and number of resource units dependent
on bandwidth.
resource unit
Number of available resource
3 6 12 24 36 48
units
Using other values such as, e.g. M=15 or M=12 or M=10 or M equal to other values can be considered
based on the outcome of the interference coordination study.
One or more RUs can be assigned to a UE by the Node B. When more than one LRUs are assigned to a
UE, they should be contiguous in frequency domain. When more than one DRUs are assigned to a UE,
the sub-carriers assigned should be equally spaced. The multiplexing of localized and distributed
transmission is FFS.
The information required by the UE to correctly identify its resources assigned must be made available
to the UE by the scheduler. The detailed signalling support is FFS.
- By transmitting each uplink reference signal across a distinct set of sub-carriers, as in “Figure
9.1.1.2.2-2 left.” This solution achieves “signal orthogonality in the frequency domain” and
applies to both localized and distributed reference-signal structures. This approach is referred
to as FDM below.
- By constructing reference signals that are orthogonal in the “code domain”, with the signals
transmitted across a common set of sub-carriers (example with contiguous sub carriers in
Figure 9.1.1.2.2-2 right). As an example, individual reference signals may be distinguished by
a specific cyclic shift of a single CAZAC sequence. This approach is referred to as CDM
below.
- Orthogonality in the time domain
- A combination of the methods above
Release 7 76 3GPP TR 25.814 V7.1.0 (2006-09)
Figure 9.1.1.2.2-2 Reference-signal orthogonality in frequency domain (left) and “code” domain
(right) respectively.
Note that orthogonality in the frequency domain is also possible for a localized reference-signal
structure.
The applicability of different reference-signal structures to different transmission structures is as
follows.
- To multiplex reference-signals from different UEs occupying different data spectrum, FDM is
used.
- Localized reference signal occupying the same spectrum as data transmission or
Distributed reference signal confined within the same bandwidth as the data transmission but
occupying a fraction of the data spectrum can be used.
- Multiplexing of reference signals for the case of a UE with multiple antennas or multiple UEs
in MU-MIMO is to be studied further.
Reference-signal for demodulation/detection in case of distributed data transmission:
- Reference signal distributed to allow for channel estimation of the distributed data. As
mentioned, for FDM, due to the use of short blocks (SB) for reference-signal transmission,
each reference-signal “comb finger” is twice as large as the corresponding “comb-finger” for
the distributed data transmissions in the long blocks (LB). Thus to provide better frequency
sampling of the channel for the channel estimation, frequency-domain staggering of the
reference signals of SB2, relative to SB1 may be applied when both the SB1 and SB2 are used
for reference-signal for demodulation/detection of distributed data transmission as shown in
Figure 9.1.1.2.2-3. Note that the staggering of distributed reference signals in SB1 and SB2
could also be used in case of localized data transmission.
- Reference signal that occupies a set of sub-carriers, which may overlap the sub-carriers which
are used by the long block (data). An example is portrayed in Figure 9.1.1.2.2-4.
Release 7 77 3GPP TR 25.814 V7.1.0 (2006-09)
Figure 9.1.1.2.2-4 An example of overlap the sub-carriers which are used by the long block
(data).
- For multiplexing reference-signals from different UEs within the same Node B, distributed
FDM and/or CDM is used.
- Reference signal may occupy at least partly different spectrum than data transmission. This
allows for channel-quality estimation also for other frequencies than that used for data
transmission and, as a consequence, allows for uplink channel-dependent scheduling.
- For multiplexing reference-signals from different UEs within the same Node B, distributed
FDM and/or CDM is used.
- Multiplexing of reference signals for the case of a UE with multiple antennas or multiple UEs
in MU-MIMO is to be studied further.
- When reference-signal for uplink channel-quality estimation is transmitted with data symbols
within the same sub-frame, a part of this reference-signal can also be used for channel
estimation for demodulation/detection of the data symbols.
The two SBs can be used for transmission of reference-signals for different purposes listed above.
When the reference signals occupying the different size of the spectrum (FFS) are multiplexed into
overlapped frequency band, different sub-carriers should be assigned for the reference signals within
the overlapping frequency band in order to achieve orthogonal transmission.
The uplink reference signals are based on CAZAC sequences. Which exact type of CAZAC sequences
is FFS.
- data-associated signaling (e.g., transport format and HARQ information), which is associated
with uplink data transmission, and
- data-non-associated signaling (e.g., CQI and/or ACK/NACK due to downlink transmissions,
and scheduling requests for uplink transmission).
Furthermore, three multiplexing combinations for the uplink pilot, data, and L1/L2 control signaling
within a sub-frame are considered for a single UE:
Figure. 9.1.1.2.3-2 shows a multiplexing scheme for L1/L2 control signaling, data, and pilot. In Figure
9.1.1.2.3-2(a), both data-associated and data-non-associated control signaling are time-multiplexed
with data and pilot within the sub-frame. Furthermore, the data-associated and data-non-associated
control signalling from multiple UEs are multiplexed in the frequency or/and code domains associated
with multiple pilot channels. In Figure 9.1.1.2.3-2(b), the data-associated control signaling is time-
Release 7 78 3GPP TR 25.814 V7.1.0 (2006-09)
multiplexed with data similar to the case in Figure 9.1.1.2.3-2(a). The data-non-associated control
signaling can also be time-multiplexed with data if UE has UL data transmission. Meanwhile, the data-
non-associated control signaling for UEs that transmits only the L1/L2 control, is multiplexed
exclusively in a semi-statically assigned time-frequency region. The data-non-associated control
signaling of different UEs is multiplexed using the frequency/time/code domain or a hybrid of them
within the assigned time-frequency region. The exclusive time-frequency region can be separated into
two frequency-time resources. First part can contain data-non-associated control signaling without user
identification, e.g. ACK/NACK, and the second part can contain the one with user identification. The
possibility for multiplexing of data-non-associated control signaling with data channel by exclusive
frequency resource, i.e., frequency-multiplexing, is FFS.
The amount of overhead due to the L1 and L2 signaling and the exact mapping to the time-frequency
resources needs further investigation.
Note that Figure 9.1.1.2.3-2 (a) - (b) show localized allocation only but the multiplexing options
described above are also applicable to distributed allocation.
Time
Transmitted
Control and Data Add CP
Pilot
Pilot
DFT
data
IFFT
sequence
Sub-carrier mapping
(Note: This figure is used for illustration purposes only, and the positions of pilot, control, and
data within the sub-frame do not specify the actual configuration.)
1 resource unit
Frequency
1 resource unit
Frequency
Pilot Pilot Pilot Pilot
Control and Control and Control and Control and
1 sub-
Data for Data for Data for Data for
frame
UE A UE B UE C UE D
Figure 9.1.1.2.3-2 – Multiplexing scheme for L1/L2 control signaling, data, and pilot
(Note: These figures are used for illustration purposes only, and the positions of pilot, control,
and data within the sub-frame do not specify the actual configuration.)
In the case of time synchronization being present, the outband control signaling consists of
In the case of time synchronization not being present, the outband control signaling consists of
Data-associated control signaling can only be transmitted together with user data. The content is
summarized in Table 9.1.2.4.1.1-1.
Release 7 80 3GPP TR 25.814 V7.1.0 (2006-09)
If
asynchronous Redundancy 2 To support incremental redundancy.
hybrid ARQ version
is adopted
New data To handle soft buffer clearing.
1
indicator
If Retransmission 2 Used to derive redundancy version (to
synchronous sequence support incremental redundancy) and ‘new
hybrid ARQ number data indicator’ (to handle soft buffer
is adopted clearing).
FFS The uplink transport format (modulation
format, transport block size, etc). Only
TF
Transport format
required if UE-based TFC selection is
supported.
Note: It is FFS whether asynchronous or synchronous hybrid ARQ operation will be adopted.
Note: It is FFS whether the transport format the UE uses is mandated by the Node B or controlled by
the UE.
Note: In case of multi-layer transmission, multiple instances of (parts of) the data-associated control
signaling may be required.
9.1.2.4.1.2 CQI
The CQI informs the scheduler about the current channel conditions as seen by the UE. If MIMO
transmission is used, the CQI includes necessary MIMO-related feedback.
9.1.2.4.1.3 ACK/NAK
The hybrid ARQ feedback in response to downlink data transmission consists of a single ACK/NAK
bit.
The synchronized random access is used by the UE to request resources for uplink data transmission. It
is further described in Section 9.1.2.1.
To achieve high processing gain, repetition coding can be used as a complement to FEC.
The control channel is multiplexed in time domain and may preferably be mapped on the symbols from
which the CP is constructed. The control channel may be transmitted in one or more data block
(number and position are FFS)
The use of both open loop transmit diversity techniques based on block codes as well as cyclic shift
diversity, open-loop and closed-loop MIMO techniques, e.g. spatial division multiplexing (SDM) and
precoding, should be considered. For the closed-loop mode, techniques for reducing signaling overhead
should be evaluated. The possibility for single user higher-order uplink MIMO (more than two TX/RX
antennas) should be considered.
Transmit antenna selection at the UE, which assumes fewer RF chains than the number of transmit
antennas (e.g. 1 RF chain and 2 transmit antennas), should be considered to potentially lower the UE
complexity.
The possibility for SDMA should also be considered. A specific example of SDMA corresponds to a
(2x2) multi-user MIMO, where two UEs, each of which transmitting on a single antenna, share the
same time and frequency resource allocation. These UEs apply mutually orthogonal reference signal
patterns in order to simplify Cell site processing (cancellation). Note that from the UE perspective
difference between (2x2) multi-user MIMO and single antenna transmission is only the use of a
reference signal pattern allowing for “pairing” with another UE.
For example, modifications to the basic modulation schemes in section 9.1.1, such as per-symbol phase
rotations (π/4-QPSK, π/2-BPSK) and I/Q-offsetting (offset-QPSK, offset-QAM), should be considered.
Release 7 82 3GPP TR 25.814 V7.1.0 (2006-09)
For example, frequency-domain spectrum shaping can be applied between the output of the DFT and
the input of the sub-carrier mapping in Figure 9.1.1-1. The selection of the filter shape is a trade-off
between spectrum/link efficiency and power de-rating reduction. For a given spectral efficiency,
different spectrum-shaping functions can provide different power de-rating reductions. Note that CM
and PAPR are indicators of power de-rating. CM is a more appropriate indicator than PAPR of power
de-rating achieved by different power de-rating reduction techniques. The use of spectrum shaping,
including the use of different spectrum shaping parameters for different modulation schemes, number
of sub-carriers or different scenarios (e.g. capacity/bandwidth limited vs. coverage/power limited),
should be considered. Different spectrum shaping functions should be further studied and optimized
for different uplink modulation formats considered. In the case of π/2-BPSK modulated signals, both
the PAPR and the CM can be reduced significantly by spectrum shaping using the Kaiser window
without reducing the spectral efficiency.
Another candidate for power de-rating reduction is the FFT Pre-processing technique. In this approach
selected input modulation symbols are attenuated in order to reduce power de-rating as indicated by
PAPR/CM at the output of the IFFT. This FFT pre processing approach is valid for any FFT size M and
IFFT size N, and for both “localised” or “distributed” sub-carrier variants. This scheme can also be
combined with pulse shaping for example RRC filtering implemented in the frequency domain.
For co-existing LCR-TDD based frame structure (Figure 6.2.1.1-1), random access time/frequency
resources is illustrated in Fig. 9.1.2.1.1.1-2, where physical random access channel and UpPCH
channels are used for L1 random access procedure.
The minimum bandwidth, BWRA, allocated for non-synchronized random access transmission is 1.25
MHz. For system bandwidths larger than 1.25 MHz, either the random access transmission uses a
larger bandwidth, or multiple random access channels are defined. Multiple 1.25MHz random access
channels might be especially useful for selecting a best block using frequency selective channel
characteristics (TDD mode). BWRA of less than 1.25 MHz (e.g. BW of the uplink resource unit = 375
KHz) for non-synchronized access is FFS.
Release 7 83 3GPP TR 25.814 V7.1.0 (2006-09)
The length of the non-synchronized random access burst, TRA, is less than (multiples of) sub-frames
(e.g. 0.5 ms) to allow the burst, and the required guard time to account for the uplink timing uncertainty
and the propagation loss, to fit within a subframe (or multiples thereof). The random access burst
length can be adjusted (e.g. on a cell basis depending on the cell size) to optimize the overhead/latency
versus coverage requirements trade-off. It is FFS on how this adjustment is made (e.g. static, semi-
static, dynamic).
For co-existing LCR-TDD based frame structure, TRA is less than 0.8 ms for optimum coverage
(combining UpPTS (0.125ms) special timeslot and TS1 (0.675ms) timeslot) to allow the burst, and the
required guard time to account for the uplink timing uncertainty, to fit within a 0.8ms. Longer TRA may
be needed for large cells.
For the co-existing LCR-TDD based frame structure, CDM may be used on physical random access
channel (Figure 9.1.2.1.1.1-2) for transmitting control message. The effect of interference on scheduled
data and vice-versa needs to be investigated further.
Data transmission
BWRA
Guard time
TRA
TS0 TS1 TS2 TS6
UpPCH
1
PRACH 1
UpPCH
Data 2 Data Data
Transmission
GP1 Transmission Transmission
UpPCH Data
3 Transmission
UpPCH
4
5ms
Figure 9.1.2.1.1.1-2 TDM/FDM option example for coexisting LCR-based frame structure (TDD
mode) for optimum coverage scenario and preamble-only transmission in the UpPCH channel
Figure 9.1.2.1.1.1-3 Example of CDM of Scheduled Channels and random access preamble
The non-synchronized random access preamble is used for time alignment, signature detection etc. The
message payload may comprise any additional associated signaling information, e.g., a random ID. A
message payload (if any) of 4-8 bits is implicitly transmitted in the random access burst along with
the preamble as shown in Figure 9.1.2.1.1.2-1. It is FFS on how the message (if any) is implicitly sent
with the preamble burst.,
Preamble part
Prior to attempting a non-synchronized random access, the UE shall synchronize to the downlink
transmission.
• Approach#1: Figure 9.1.2.1.1.3-1 outlines this approach, where the Node B responds to the
non-synchronized random access attempt with timing information to adjust the uplink
transmission timing and an assignment of uplink resources to be used for transmission of data
or control signalling (possibly including any message payload (e.g. UE ID) not included in the
preamble) using the shared data channel. It may be noted that the timing information can also
be combined with the uplink data resource allocation. Furthermore, the uplink data resource
allocation may be implicitly indicated by associating a reserved time frequency region with a
preamble sequence.
• Approach#2: Figure 9.1.2.1.1.3-2 outlines this approach, where the Node B responds to the
non-synchronized random access attempt preamble with timing information and resource
allocation for transmission of scheduling request (and possibly any additional control
signalling or data). UE then sends the scheduling request at the assigned time-frequency
resource using the shared data channel or physical random access channel (for co-existing
LCR-TDD based frame structure). The Node B adjusts the resource allocation according to
the scheduling request from the UE.
UE Node-B
Timing information
UL Data Transmission
The power control scheme shall be designed assuming no intra-cell interference from data
transmissions (i.e., TDM/FDM operation).
Open loop power control is used to determine the initial transmit power level. It is possible to vary the
random access burst transmit power between successive bursts using:
a) Power ramping with configurable step size including zero step size for both FDD and TDD case
Synchronized random access and data transmission are also time and/or frequency multiplexed. An
example of synchronized random access (applicable for FDD, TDD and co-existing LCR TDD) is
shown in Figure 9.1.2.1.2.1-1. It may be noted that the number of long blocks (LB) shown in the
figure is for illustration only. The minimum bandwidth, BWRA, allocated for synchronized random
access transmission is equal to the bandwidth of the uplink resource allocation unit (e.g. 375 KHz).
Also, the random access transmission can use a larger bandwidth, or multiple random access channels
using the minimum bandwidth can be defined. Multiple random access channels might be useful for
selecting a best block using frequency selective channel characteristics (TDD mode). The length of
the synchronized random access burst can be adjusted (e.g. on a cell basis depending on the cell size) to
Release 7 87 3GPP TR 25.814 V7.1.0 (2006-09)
optimize the overhead/latency versus coverage trade-off.. It is FFS on how this adjustment is made
(e.g. static, semi-static, dynamic). Synchronized random access can be done every x sub-frames (e.g.
x=2).
Data transmission
BWRA
1 DFT-SOFDM Symbol
1 subframe
For synchronized random access, Figure 9.1.2.1.1.3-1 and 9.1.2.1.1.3-2 also apply, except the timing
information may not be transmitted.
The preamble sequences shall be designed assuming no intra-cell interference from data transmissions
(i.e., TDM/FDM operation).
The random access channel sequence(s) (e.g. based on CAZAC/GCL) used to generate the transmitted
random access preamble waveforms should have the following properties:
1. Good detection probability while maintaining low false alarm rate e.g. by maximizing post-
decoder Es/(Nt+Ne) for a occupied random access channel preamble where Ne is the residual
interference due to other random access channel transmissions in a given random access
channel and Nt is thermal noise.
2. Number of random access channel preamble waveforms should be defined to handle the
maximum expected multiple access scenarios (traffic load) while guaranteeing low collision
probability.
3. Enable accurate timing estimation (e.g. good autocorrelation properties and sufficient
occupied BW).
4. Low power de-rating (low CM/PAPR).
9.1.2.2 Scheduling
The uplink should allow for both scheduled (Node B controlled) access and contention-based access.
In case of scheduled access the UE is dynamically allocated a certain frequency resource for a certain
time (i.e. a time/frequency resource) for uplink data transmission. Downlink control signaling informs
UE(s) what resources and respective transmission formats have been allocated. The decision of which
user transmissions to multiplex within a given sub-frame may for example be based on
Release 7 88 3GPP TR 25.814 V7.1.0 (2006-09)
Methods to reduce the control signaling overhead, e.g., pre-configuring the scheduling instants
(persistent scheduling) and grouping for conversational services, should be considered. Transmission of
the reference signals to facilitate the uplink channel quality measurements should be investigated. In
addition it should be determined if grouping can more efficiently use time frequency resources
resulting in higher capacity.
However, some time/frequency resources can be allocated for contention-based access. Within these
time/frequency resources, UEs can transmit without first being scheduled. As a minimum, contention-
based access should be used for random-access and for request-to-be-scheduled signaling
In unpaired spectrum, system capacity may be improved through the use of localised FDMA
contention-based access channels. The UE may select the access channel based upon knowledge of the
channel state information measured on a recent downlink sub-frame
For this purpose, uplink link adaptation should effectively utilize a combination of the adaptive
transmission bandwidth accompanied with channel-dependent scheduling, transmission power control,
and the adaptive modulation and channel coding rate.
Three types of link adaptation are performed according to the channel conditions, the UE capability
such as the maximum transmission power and maximum transmission bandwidth etc., and the required
QoS such as the data rate, latency, and packet error rate etc. In particular, the three schemes are
controlled by channel variation as link adaptation. The basic features of the three link adaptation
methods are as follows.
• The same coding and modulation is applied to all resource units assigned to which the
same L2 PDU is mapped on the shared data channel scheduled for a user within a TTI.
This applies to both localized and distributed transmission. The overall coding and
modulation is illustrated in Figure 9.1.2.3-1.
Transport block (L2 PDU)
CRC attachment
Channel coding
HARQ functionality
including adaptive
coding rate
Physical channel
segmentation Number of assigned
(resource unit mapping) resource units
Adaptive modulation
(common modulation is selected)
The control update interval for each of the three link adaptation methods should be jointly investigated
from the viewpoint of the achievable performance such as the throughput, packet error rate, and latency
as well as the required signaling overhead.
If the uplink HARQ operation is synchronous in time, two different kinds of link adaptation techniques
may be considered:
- The associated UL assignment is sent by the Node-B for the first transmission and all
subsequent retransmissions
• The Node-B has the flexibility to change the modulation order, the set of coded rate-
matched bits and resource unit allocation in an adaptive manner
• For the retransmissions, the Node-B only needs to signal the transmission attributes
which are adapted on a non-predetermined basis
- The associated UL assignment is sent by the Node-B only for the first transmission
• The number of allocated resource units is fixed.
• The sequence of resource unit mapping, modulation order and set of coded rate-
matched bits for each retransmission is pre-determined and known to the Node-B and
UE.
The impact of such schemes on link and system performance needs to be evaluated further.
Release 7 90 3GPP TR 25.814 V7.1.0 (2006-09)
The slow power control may be implemented in each Node-B by sending slow updating power control
signaling. Alternatively, each UE can derive its own transmission power according to the path loss
measurement from downlink pilot.
To achieve good trade-off of the cell-edge performance and the overall spectral efficiency, slow power
control scheme that compensates a fraction of the path loss and shadowing should be considered.
9.1.2.5 HARQ
Uplink HARQ should be based on Incremental Redundancy. Note that Chase Combining is a special
case of Incremental Redundancy and is thus implicitly supported as well.
- Synchronous HARQ implies that (re)transmissions for a certain HARQ process are restricted
to occur at known time instants. No explicit signaling of the HARQ process number is
required as the process number can be derived from, e.g., the subframe number.
- Asynchronous HARQ implies that (re)transmission for a certain HARQ process may occur at
any time. Explicit signaling of the HARQ process number is therefore required.
In principle, synchronous operation with an arbitrary number of simultaneous active processes at a time
instant could be envisioned. In this case, additional signaling may be required. Asynchronous operation
already supports an arbitrary number of simultaneous active processes at a time instant. Furthermore,
note that, in a synchronous scheme, the transmitter may choose not to utilize all possible retransmission
instants, e.g., to support pre-emption. This may require additional signaling.
The various forms of HARQ schemes are further classified as adaptive or non-adaptive in terms of
transmission attributes, e.g., the Resource unit (RU) allocation, Modulation and transport block size,
and duration of the retransmission. Control channel requirements are described for each case.
- Adaptive implies the transmitter may change some or all of the transmission attributes used in
each retransmission as compared to the initial transmissions (e.g. due to changes in the radio
conditions). Hence, the associated control information needs to be transmitted with the
retransmission. The changes considered are:
• Modulation
• Resource Unit allocation
• Duration of transmission
- Non-Adaptive implies that changes, if any, in the transmission attributes for the
retransmissions, are known to both the transmitter and receiver at the time of the initial
transmission. Hence, the associated control information need not be transmitted for the
retransmission.
Release 7 91 3GPP TR 25.814 V7.1.0 (2006-09)
With those definitions, the HS-DSCH in WCDMA uses an adaptive, asynchronous HARQ scheme,
while E-DCH in WCDMA uses a synchronous, non-adaptive HARQ scheme.
The capability of adaptively being able to change the packet format (i.e., adaptive IR) and the
transmission timing (i.e., asynchronous IR) yields an adaptive, asynchronous IR based HARQ
operation. Such a scheme has the potential of optimally allocating the retransmission resources in a
time varying channel. For each HARQ retransmission, control information about the packet format
needs to be transmitted together with the data sub-packet.
Synchronous HARQ transmission entails operating the system on the basis of a predefined sequence of
retransmission packet format and timing.
The benefits of synchronous HARQ operation when compared to asynchronous HARQ operation are:
- Reduction of control signalling overhead. from not signalling HARQ channel process number
- Lower operational complexity if non-adaptive operation is chosen
- Possibility to soft combine control signalling information across retransmissions for enhanced
decoding performance if non-adaptive operation is chosen.
Therefore, for the purpose of the feasibility study, synchronous HARQ operation is assumed for the
SC-FDMA based E-UTRA uplink. The impact of ACK/NAK signalling errors on synchronous HARQ
operation needs further study.
Depending on the actual L1/L2 requirements, asynchronous HARQ may best address the issues of:
- Scheduling flexibility if fully adaptive operation is selected and if both localized and
distributed allocations are selected.
- Support for multiple simultaneous (in the same (set of) subframe(s)) independent HARQ
processes
- Flexibility in scheduling of retransmissions
The desirability of particular L1/L2 features will determine the degree of adaptive operation.
If the UE does not have uplink data to transmit for a longer period, no uplink transmission should be
carried out. In that case, uplink time alignment may be lost and restart of data transmission must then
be preceded by an explicit timing-re-alignment phase to restore the uplink time alignment.
- Inter-cell-interference cancellation
- Frequency domain spreading
Regarding the Frequency domain spreading, a spreading gain can be obtained either explicitly by
spreading modulation symbols over multiple carriers or implicitly by using repetition code in the
channel coding.
The slow power control (9.1.2.4.1) and power control based upon neighbour cell load (9.1.2.4.2) can
also be seen as a means for uplink inter-cell-interference mitigation.
In addition, the use of beam-forming antenna solutions at the base station is a general method that can
also be seen as a means for uplink inter-cell-interference mitigation.
It should be noted that the different approaches could, at least to some extent, complement each other
i.e. they are not necessarily mutually exclusive.
Regarding required inter-cell interference co-ordination in support for uplink interference co-
ordination, two cases are considered similar to downlink interference co-ordination:
A pseudo-random method can be used to generate the user/cell-specific interleaver patterns for IDMA.
The number of the available patterns (seeds) is determined by the length of interleaver. A Node B can
identify the interleaver pattern of the cell by checking its interleaver pattern ID. The seeds can be
reused between “far-spaced” cells in a manner similar to that of frequency reuse in a cellular system.
The benefits from the randomization of the inter-cell interference come without any restriction on the
Node B scheduler or receiver type. However, these benefits are achievable only if the transmitted
waveform characteristics support the means to randomize the inter-cell interference. Therefore, means
to randomize the inter-cell interference experienced in the reception of all the UE signals may be
supported by the definition of the UL signal characteristics.
The IDMA based inter-cell-interference cancellation scheme would imply the following requirements
on the system:
- Band allocation: The overlapped frequency resource in the “cell edge” area should be reused
with the same “band allocation” in the serving and interfering cells respectively. And the
“interfering UE” and the “interfered UE” should transmit using the same band.
- Synchronization: Inter-NodeB synchronization is required.
- Intra-cell signalling: When IDMA is used, a NodeB naturally has the knowledge of the
interleaver pattern used by the UEs in the cell, hence no extra signalling is needed.
- Inter-cell signalling: Interfering UE configurations (e.g. interleaver pattern ID, modulation
scheme, FEC scheme and coding rate) should also be signalled to the interfered NodeB. To
cancel the inter-sector interference in uplink, the NodeB naturally has the knowledge of
interfering UE, hence no extra signalling or operation is needed. To cancel the inter-NodeB
interference, the signalling of interfering UE configurations to the NodeB is realized by
detecting the control channel of the interfering UE.
It may be noted that the specified numerology is for evaluation purpose only.
The minimum TTI for uplink transmission is equal to the uplink sub-frame duration. Similar to the
downlink, the possibility to concatenate multiple sub-frames into longer uplink TTIs should be
considered.
Release 7 94 3GPP TR 25.814 V7.1.0 (2006-09)
Note that the sub-carrier spacing is constant regardless of the transmission bandwidth. To allow for
operation in differently sized spectrum allocations, the transmission bandwidth is instead varied by
varying the number of OFDM sub-carriers.
1.) In band pilots - used for coherent data demodulation, e.g. channel estimation. These pilots are
transmitted in the part of the bandwidth used for data transmission.
2.) Out of band pilots – used for advanced frequency dependent scheduling and link adaptation.
These pilots span a larger bandwidth than the one used for data transmission.
Note that in band pilots may also be used for frequency dependant scheduling and link adaptation.
• If a UE transmits on two antennas (Ant A and Ant B) as in the case of MIMO or Tx diversity
• If multiple UEs share the same time and frequency resource, each of the UEs transmitting on a
single antenna it is beneficial that UEs use orthogonal pilot patterns (this is described as multi-
user MIMO, a specific case of SDMA, see section 9.2.1.3).
Orthogonality of in-band pilot symbol patterns can be achieved in the time and/or frequency domain.
Figure 9.2.1.2-1 shows an example of the IBP locations and overheads in the case channel allocation to
a UE in the time domain is done in multiple of 7 symbols (a full sub-frame). The exact pilot locations
and overhead are FFS.
Figs-9.2.12-2 (a) – (c )show examples of the IBP location and overheads in the case channel allocation
in the time domain to a UE is done in multiple of 6 symbols, meaning that the first symbol in a sub-
frame may be used for other purposes (e.g. common control signalling). The exact pilot locations and
overhead are FFS. Figs-9.2.12-2 (a) – (c) further exemplify different cases of pilot pattern
orthogonality:
• Fig 9.2.1.2-2 (a) exemplifies the case of a single UE transmitting on a single antenna for which
no orthogonal pilot is used.
• Fig 9.2.1.2-2 (b) exemplifies the case of a UE transmitting on multiple antennas for which
orthogonal pilot patterns are transmitted from the multiple antennas
• Fig 9.2.1.2-2 (c) exemplifies the case of multiple UEs, each of which transmitting on a single
antenna, sharing the same time and frequency resource. Each UE transmits one orthogonal pilot
pattern (multi-user MIMO case)
Release 7 95 3GPP TR 25.814 V7.1.0 (2006-09)
7 OFDM symbols
0.5ms
Data symbols
10 sub-carriers
6 OFDM
In-band Pilot
Data symbols
10 sub-carriers
Figure 9.2.1.2-2 (a): Channel allocation in the time domain in multiple of 6 symbols
UE (Ant A) UE (Ant B)
Null
6 OFDM
In-band Pilot
Data
10 sub-carriers
Figure 9.2.1.2-2 (b): Channel allocation in the time domain in multiple of 6 symbols
6 OFDM symbols
In-band pilots (IBP)
Data symbols
Null
10 sub-carriers
Figure 9.2.1.2-2 (c): Channel allocation in the time domain in multiple of 6 symbols
9.2.1.3 MIMO
The baseline antenna configuration for uplink single-user MIMO is two transmit antennas at the UE
and two receive antennas at the Cell site.
The possibility for single-user higher-order uplink MIMO (more than two TX/RX antennas) should be
considered.
The possibility for SDMA should also be considered. A specific example of SDMA corresponds to a
(2x2) multi-user MIMO, where two UEs, each of which transmitting on a single antenna, share the
same time and frequency resource allocation. These UEs apply mutually orthogonal pilot patterns in
order to simplify Cell site processing (cancellation). Note that from the UE perspective difference
between (2x2) multi-user MIMO and single antenna transmission is only the use of a pilot pattern
allowing for “pairing” with another UE.
Assuming total of N available tones and K tones are reserved. Let X be frequency-domain data signal
and C = [C0, C1, …, CK-1] be a code on subset R. The goal of TR method is to find the optimum code
value C so that:
ˆ & || ∞ <|| [ || ∞
min || [ + F || ∞ = min || [ + 4 (9.2.1-1)
.
In [x], a simple gradient algorithm with fast convergence is proposed. The overall TR iterative
algorithm is simply:
xi +1 = xi − µ ⋅ ∑α p i
n n (9.2.1-2)
xni >A
n is the index for which sample xn is greater than the clipping threshold;
and p n is called peak reduction kernel vector. The kernel is a time domain signal that is as close as
possible to the ideal impulse at the location where the sample amplitude is greater than the predefined
threshold. This way the peak could be cancelled as much as possible without generating secondary
peaks. p n is derived from original kernel p 0 through right circle shifting (by n-1 samples). The
original kernel p 0 can be calculated using 2-norm criteria and is given by the following formula:
N ˆ
p0 = Q1K (9.2.1-4)
K
1K is a vector of length K with all one elements.
In a example of the improved tone reservation with reduced complexity all tones except guard band [y]
are used to calculate an original kernel. Then, α combined with µ is quantified to form derived
reduction kernels. The phase is divided equally into s parts. And the amplitude is divided into t parts
represented by some special values according to different FFT size and step length. For example, if
FFT size is 1024, the phase is divided equally into six parts represented by ± π 6, ± π 2, ± 5π 6
and the amplitude can be chosen among 0.01, 0.04, 0.08, 0.12, 0.16. Thus only 30 peak reduction
kernels need to be stored.
In order to reduce the computation load, only choose fixed number of peaks to be cancelled in one
iteration instead of all the peaks that satisfies xni > A .
The Steps of the improved TR method with reduced complexity is described below:
If all samples are below the target threshold, transmit x . Otherwise, search among
i
4.
the derived kernels (stored in advance) to find matched ones according to Equation 3
and right circle shift them by ni samples;
Update x according to Equation 2;
i
5.
Repeat step 3 to step 5 until i reaches maximum iteration limit. Transmit final x .
i
6.
9.2.1.4.2 Circulated clipping and filtering
Clipping is a simple method for PAPR reduction. In order to reduce PAPR meanwhile keep the
spectrum characters of the signal, circulated clipping and filtering can be used [1].
Letx(n) stands for transmitted signals without clipping, y (n) stands for signals after clipping, and
A is the threshold for clipping which is related to clipping ratio. One of clipping criterions is as the
following equation.
A 2
x ( n), if x (n) > A2
2 (9.2.1-5)
y ( n) = | x( n) |
2
x ( n), if x (n) ≤ A
2
To suppress out-of-band leakage caused by clipping, filtering has to be added. Since filtering causes re-
growth to PAPR, clipping and filtering are repeated in circles for times to depress the PAPR,
meanwhile reduce out-of-band leakage to an acceptable degree. The principle of circulated clipping and
filtering is given in Figure 9.2.1.4.2-1.
• Frequency reuse 1
• Node-B scheduling
The following additions to the baseline multiple access structure should be considered:
• Enhanced uplink control structure and procedures in support of HS-DSCH and E-DCH
operation with variable symmetric and asymmetric bandwidth allocations.
The system operation should rely on the definition of new E-DPDCH demodulation performance
requirements based on the following Node B receiver techniques:
The concepts presented are valid for multi-carrier operation based on the 5MHz system (UTRA FDD)
as well as the 1.25MHz system (WCDMA LCR) or a 5MHz/1.25MHz hybrid multi-carrier system.
9.3.2.1 Definitions
Refer to section 7.2.2.1 for definitions relevant for EUTRA operation based on MC-WCDMA multi-
carrier operation.
The timing of the PHY channels for paired carriers shall be no different than for a single carrier
system where the timing of the UL channels is always referenced to the timing of associated DL
channels (see 25.211 for a complete reference).
The timing of the PHY channels for unpaired carriers is explicitly covered in this Technical Report.
One cell is the serving E-DCH for all carriers supported by a given UE.
HARQ PHY re-transmissions on UL takes place at the same carrier as for the first
transmission.
Figure 9.3.2.2-1 is a block diagram depicting multi-carrier operation. Each of the colors represents a
different UL carrier. Note that the PHY channels in squared brackets are just transmitted if associated
downlink carrier is configured.
Release 7 100 3GPP TR 25.814 V7.1.0 (2006-09)
Node-B 1 * DPCCH
* E-DPCCH
* E-DPDCH
[* HS-DPCCH]
[* DPDCH single carrier]
Controller
Node-B 3
UE
Node-B 2
From the UE viewpoint, the system is accessed by way of the anchor carrier. In turn, the UE shall
expect reception of the corresponding AICH (Access Indicator Channel) at the carrier associated with
the one used for transmission of the PRACH. For DCH transmission, the UE is expected to use at most
one carrier. Multi-carrier transmission is limited to the E-DCH.
E-DPDCH M
M=N: All the UL carriers have an associated DL carrier and vice-versa. PHY procedures for
this case (i.e., Power Control, synchronization, HS-DSCH and E-DCH related procedures…)
are no different than those for the single carrier case.
M>N: Just the N paired carriers will carry the HS-DPCCH and the TPC commands for the N
downlink carriers. Therefore, there will be (M-N) uplink DPCCH with no need for DL power
control TPC commands.
M<N: There are (N-M) downlink unpaired carriers. Therefore, besides the M paired carriers
carrying HS-DPCCHs, the ACK/NACK and CQI information of (N-M) unpaired DL carriers
will have to be conveyed from the UE to the E-UTRAN in some way. All the M downlink (F-
)DPCHs are power controlled by the paired carriers. The proposed method to convey this
additional control information is presented below.
To convey the HSDPA feedback information (i.e., ACK/NAK channel and CQI channel) of the (N-M)
unpaired DL carriers is to code division multiplex (N-M) additional HS-DPCCHs within the M uplink
carriers.
Doing this code division multiplexing requires the definition of the channelization code to be used by
the additional HS-DPCCHs within a carrier. For the single carrier system, 25.213 specifies the SF 256
channelization code and the quadrature phase (depending on the number of DPDCHs configured) to be
used by the only HS-DPCCH that may be transmitted at a UE. Therefore, 25.213 would have to further
define the channelization code and the quadrature phase to be used by the additional HS-DPCCHs.
The additional HS-DPCCHs themselves would be no different to the HS-DPCCHs of the paired
frequencies or the HS-DPCCH of the current single carrier system. The timing of these additional
channels would be tied to the associated downlink HS-PDSCH.
As a general rule, in order to limit the impact on peak-to-average of the transmit waveform with an
additional code channel, the (N-M) additional HS-DPCCHs would have to be spread across the M
uplink carriers as much as possible.
In a similar way as Table 0 in 25.213, the following table summarizes the maximum number of
simultaneous uplink dedicated PHY channels configurable at a given carrier.
1:1 1 1 1 4
1:2 1 2 1 3
1:3 1 3 1 2
1:4 1 4 1 1
Release 7 102 3GPP TR 25.814 V7.1.0 (2006-09)
As seen in Table 9.3.2.4.1-1, the maximum number of code channels allowed within one carrier is no
more than the current maximum number of code channels allowed in Table 0 of 25.213.
Observations:
There is one HS-DPCCH code channel for each HS-DPCCH corresponding to a downlink
unpaired carrier.
o First row in the table above is no different to the no-DPDCH configuration in Table 0
of 25.213.
Note that the HS-PDSCHs are not power controlled. The TPC bits in the uplink DPCCH
power control the downlink (F-)DPCH. Power control of the HS-SCCH and the DL E-
channels may be based on the CQI reports by the UE for each of the DL carriers.
This section covers the timing specifics for asymmetric, i.e., N ≠ M multi-carrier configurations.
In the N>M case, there are (N-M) downlink unpaired carriers. The timing corresponding to the (N-M)
additional HS-DPCCHs in the uplink is referenced to the timing of the associated downlink HS-DPCHs
and therefore, it is well defined.
In the N<M case, the timing of the PHY channels in the (M-N) unpaired uplink carriers i.e., DPCCH
and E-DPCCH timing, is well defined as it is referenced to (M-N) additional (F-)DPCH allocated
within the N downlink carriers. Note that for this case, the timing of each of the unpaired UL carriers is
referenced to one of the N downlink carriers (the one with the associated (F-)DPCH).
As indicated in section 7.2.2.5, for each of the possible N/M relative values:
N=M: each DL carrier has its associated UL carrier and vice-versa. Therefore, the N
(F-)DPCH channels will power control the N uplink DPCCHs.
N>M: there are M necessary (F-)DPCHs that power control the M uplink DPCCHs. Just the
paired carriers carry the UL TPC commands over the respective F-DPCH.
N<M: the (F-)DPCHs in the N paired carriers will power control the uplink DPCCHs in those
carriers. In addition, each cell in the UE’s active set needs to allocate (M-N) additional (F-
)DPCHs within the N downlink carriers to perform power control of the uplink DPCCH of the
(M-N) uplink unpaired carriers.
M=N: each UL carrier has its associated DL carrier and vice-versa. Therefore, the M
downlink E-HICH, E-RGCH and optionally E-AGCH will control the corresponding M unlink
E-DPCHs.
M>N: the E-HICH, E-RGCH and optionally E-AGCH in the N paired carriers will control the
uplink E-DPCHs in those carriers. In addition, each cell in the UE’s active set needs to
allocate (M-N) additional E-HICH, E-RGCH and optionally E-AGCH within the N downlink
carriers to control the uplink E-DPCH of the (M-N) uplink unpaired carriers. How that
additional information is conveyed is subject to different options covered in section 7.2.2.5.
M<N: the E-HICH, E-RGCH and optionally E-AGCH in the M paired carriers will control the
uplink E-DPCHs in those carriers.
9.3.3.4.1 E-DPCH Retransmission on Multi-Carrier system
Operation in the multi-carrier system shall guarantee PHY HARQ retransmissions on the carrier that
was used for the first transmission.
For the DL multiple access, the bandwidth of each downlink sub-carrier will be allocated as 1.6 MHz.
The baseline antenna configuration for uplink MIMO is two transmit antennas at the UE, and two
receive antennas at the Cell site. The possibility for more receive antennas should also be considered.
The antenna configuration for uplink transmit diversity (beamforming) is one or two transmit antennas
at the UE, and the number of receive antennas from four to eight at the Cell site.
9.4.2.2 Scheduling
Using AMC to adjust the modulation and coding rate, adaptive link technologies improve the
performance of system.
The open-loop and close-loop power control are supported against deep fading, eliminating near-far
effect, and fighting multiple access interference.
9.4.2.5 HARQ
Incremental Redundancy (IR) should be used for uplink HARQ. Note that Chase combining is a special
case of IR
• WCDMA Release-6
• Rake receiver
Release 7 105 3GPP TR 25.814 V7.1.0 (2006-09)
Figure 10.1.1.3.1-1 presents achievable throughput as a function of the received Es/N0 per antenna
branch assuming a 1x2 configuration (i.e. single stream, 2 receive antennas) configuration for various
modulation and code rates.
50
QPSK R = 1/3
QPSK R = 1/2
40
Throughput (Mbps)
QPSK R = 3/4
16QAM R = 1/2
16QAM R = 3/4
30 16QAM R = 4/5
16QAM R = 8/9
20
10
0
-5 0 5 10 15 20 25
Average received Es/N0 per receiver branch (dB)
18
16
14
Throughput in MBPS
12
10
6
16QAM r7/8, 16QAM r1/2 MIMO
4
16QAM r7/8, 16QAM r1/3 MIMO
16QAM r7/8, 16QAM r3/4 MIMO
2
16QAM r7/8 SIMO
0
10 12 14 16 18 20 22 24 26 28
SNR (dB)
The case in Table 10.1.1.3.2-2 is the cell edge case (taken from the same simulation run by looking the
5% CDF as in Table 10.1.1.3.2-1) with the same proportional fair scheduler and interference
coordination over the 10 MHz bandwidth. The WCDMA reference gain obtained with 5 MHz
bandwidth from [3GPP TR 25.896] with no additional penetration loss but higher inter-site distance
and scaled for bits/MHz/cell for comparison (lack of the 20 dB penetration loss with WCDMA will
more than compensate the longer inter-site distance in the reference case).
Release 7 108 3GPP TR 25.814 V7.1.0 (2006-09)
Table 10.1.1.3.2-2: Full buffer, 10 MHz, 2Rx, 5% CDF Capacity (cell edge)
Reference SC-FDMA
Speed % w.r.t
Case (WCDMA 2 Rx - /MF) 2 Rx
[km/h] reference
[b/s/Hz]] [b/s/Hz]
3 3 0.026 0.087 + 234%
A summary of the uplink LTE 25.913 reference system evaluation baseline results of cell and user
throughput performance relative to the 25.913 WCDMA reference for deployment cases 1 - 3 are given
in Table 10.1.1.3.2-4. Table 10.1.1.3.2-3 indicates some key simulation assumptions used by each
source. Table 10.1.1.3.2-5 shows the throughput performance comparison relative to 25.913
WCDMA reference for the 2800 meter inter-site distance case with 0dB penetration loss using a
Pedestrian B channel at 3km/h (subsequently referred to as case 6). Note, TR 25.913 sections 7.1 and
7.2 give the relative cell and user throughput targets and the LTE and WCDMA reference UE and
Node-B configurations..
Table 10.1.1.3.2-4 – Uplink 25.913 Reference System Evaluation Baseline Results for Sources (2, 4, 6, 7)
UL Simulation 25.913 WCDMA 25.913 E-UTRA Gain ( x WCDMA Rel-6 )
Cases and T-put (2) (4) (7) (6) (2) (4) (7) (6) (2) (4) (7) (6)
Metric Type [b/s/Hz] [b/s/Hz] [b/s/Hz] [b/s/Hz] [b/s/Hz] [b/s/Hz] [b/s/Hz] [b/s/Hz]
Case 1: Sector 0.200 0.170 0.330 0.460 0.420 0.64 0.694 0.890 2.1 x 3.8 x 2.1 x 1.9 x
Case 2: Sector 0.170 - 0.320 0.380 0.470 - 0.694 0.750 2.8 x 2.2 x 2.0 x
Case 3: Sector 0.210 0.100 0.320 0.330 0.440 0.30 0.630 0.690 2.1 x 3.0 x 2.0 x 2.1 x
Case 1: AvgUser 0.020 0.130 0.033 0.091 0.042 0.62 0.069 0.177 2.1 x 4.8 x 2.1 x 2.0 x
Case 2: AvgUser 0.017 - 0.032 0.076 0.047 - 0.069 0.149 2.8 x 2.2 x 2.0 x
Case 3: AvgUser 0.021 0.085 0.032 0.065 0.044 0.28 0.063 0.138 2.1 x 3.3 x 2.0 x 2.1 x
Case 1: 5%User 0.0051 0.075 0.009 0.012 0.010 0.15 0.017 0.038 2.0 x 2.0 x 2.0 x 3.3 x
Case 2: 5%User 0.0047 - 0.009 0.017 0.010 - 0.017 0.050 2.1 x 2.0 x 2.9 x
Case 3: 5%User 0.00024 0.002 0 0.003 0.001 0.01 0 0.012 5.0 x 3.5 x 3.9 x
Table 10.1.1.3.2-5 – Uplink 25.913 Reference System Evaluation Baseline Results for Source (1)
Uplink 25.913 reference system evaluation baseline performance results given in Table 10.1.1.3.2-4 and 10.1.1.3.2-5
show that the E-UTRA uplink based on SC-FDMA has 2x to 3x the cell and user throughput of the 25.913 WCDMA
reference case.
3GPP
Release 7 110 3GPP TR 25.814 V7.1.0 (2006-09)
(1) Semi static scheduling for full rate frames; individual scheduling for SID frames. The corresponding
associated downlink control overhead has been evaluated to be in the order of 0.5 grant per TTI.
(2) Semi static UL resource allocation to each group of VoIP users. Per user UL ACK/NACK bit (sent on
DL) re-used (together with other ACK/NACK bits) to indicate MCS and partitioning of group
resource between scheduled users and when to transmit voice packets. Delay sensitive scheduling.
Table 10.1.1.3.3-2 Uplink VoIP capacity & IOT for E-UTRA (AMR 12.2 Kbps, 5% outage)
Table 10.1.1.3.3-3 Uplink VoIP capacity & IOT for E-UTRA (AMR 7.95 kbps, 5% outage)
Figure 10.1.1.3.3-1 shows one example of the distribution of the 98% delay; this corresponds to the set of results (1).
Figure 10.1.1.3.3-2 shows example of the distribution of the 98% delay; this corresponds to the set of results (2).
1
216 UEs
0.9 228 UEs
240 UEs
0.8
0.7
0.6
CDF
0.5
0.4
0.3
0.2
0.1
0
0 10 20 30 40 50 60 70 80 90 100
98% delay [ms]
3GPP
Release 7 111 3GPP TR 25.814 V7.1.0 (2006-09)
0.9
0.8
0.7
0.6
CDF
0.5
0.4
0.2 80 UEs/sector
100 UEs/sector
0.1
120 UEs/sector
0
5 10 15 20 25 30 35 40
98% delay (ms)
Figure 10.1.1.3.3-2: 1.25 MHz bandwidth; 7.95 kbps AMR codec; Case 1 (2)
The comparison presented in Table 10.1.2.1-2 assume OFDMA based system operated in non orthogonal manner (i.e.
using spatial multiplexing) with MC-WCDMA using a successive interference cancellation receiver; both operate at
IOT = 4.5 dB. Note that user separation in case of MC-WCDMA is by means of SIC while user separation in case of
(non-orthogonal) MIMO is by means of less complex LMMSE, and can be further improved with SIC.
3GPP
Release 7 112 3GPP TR 25.814 V7.1.0 (2006-09)
3GPP
Release 7 113 3GPP TR 25.814 V7.1.0 (2006-09)
Table 10.1.2.3-2 presents similar results for the SC-FDMA system based on cases 1 and 2 defined in Table A.2.1.1-1
except that the ITU channel model PA, and VA were used, respectively. The results are based on full buffer traffic
models and the same two multi-user MIMO scheduling algorithms as in the OFDMA case. SC-FDMA based system
operated in orthogonal manner is compared with systems operated in non orthogonal manner (i.e. using spatial
multiplexing). These results assume proportional fair scheduler, real channel estimation, one sounding sub-carrier per
Resource Block, and no channel sounding error.
The total delay between the HARQ entities is of the order of:
3GPP
Release 7 114 3GPP TR 25.814 V7.1.0 (2006-09)
Total delay due to ROHC, Ciphering, and RLC/MAC processing is ~ 0.5ms. The AGW-E Node-B transfer delay (Ts1) is
not included.”
The EUTRA system will provide significantly higher data rates than Release 6 WCDMA and, as a consequence hereof,
the physical layer complexity will increase accordingly compared to lower-rate systems. The increase in data rate is
achieved through higher transmission bandwidth and/or support for MIMO and will have complexity impacts such as:
This complexity is not seen as EUTRA specific, but is similar to the complexity experienced in any high data rate
system.
One of the advantages of an OFDM/SC-FDMA based system is that it allows for implementation of a lower complexity
receiver at wider bandwidths. Another advantage is the possibility of operating MBMS in a single frequency network
manner where significant performance gains can be achieved with no additional complexity increase in the UE receiver
(see section 8.1.3).
EUTRA will support multiple bandwidth options ranging from 1.25 to 20 MHz and both FDD and TDD modes. The
variable bandwidth options has a complexity impact, however with proper channel structures, e.g., designing control
channels such that decoding is invariant to the transmission bandwidth, limited additional complexity due to the
multiple bandwidths has been identified. Support for both FDD and TDD modes is not expected to have major
complexity impact provided that maximum commonality between the two modes is maintained.
Based on what was seen as acceptable increase in complexity, it has been decided that all UEs shall have a reception-
and transmission-bandwidth capability of at least 10 MHz. Limiting the bandwidth to 10 MHz creates challenges in
mobility measurements when 10 MHz UEs are receiving data from cells with 20 MHz operating bandwidth, but these
problems are solvable with a limited complexity impact
12 UE capabilities
3GPP
Release 7 115 3GPP TR 25.814 V7.1.0 (2006-09)
3GPP
Release 7 116 3GPP TR 25.814 V7.1.0 (2006-09)
Coding for data channel and Mother code rate Turbo, LDPC
Coding for control channel and Mother code Turbo, Convolutional, other
DL Peak rates
UL Peak rates
Unrecovered power
3GPP
Release 7 117 3GPP TR 25.814 V7.1.0 (2006-09)
4 different scenarios, SCM-A to D, are considered, see Table A.1.3-1, they represent a subset of “typical” antenna
configurations and propagations scenarios.
SCM-B Urban Macro (low spread) 6-sector, 0.5λ spacing Handset, data position
Note that models SCM-C and SCM-D can also be used for evaluating laptops with two receive antennas. In this case,
one should select the channel coefficients associated with one of the two dual-polarized antennas.
The multi-antenna channel model is a tapped delay line model with covariance matrices for describing the fast fading
correlation and power distribution over transmit and receive antennas.
The total per-tap covariance matrix Rtap is obtained from the Kronecker product of the polarization covariance matrix Γ
and the Node B and UE spatial correlation matrices Α and Β, further weighted by the antenna gains at Node B and UE:
where ptap is the relative power of the tap, g NodeB ,tap is the effective antenna gain at the Node B, gUE ,tap is the
1 α 1 β
antenna gain at the UE, Α= and Β = . ⊗ denotes Kronecker multiplication.
α ∗ 1 β ∗ 1
To determine Γ unambiguously, the antenna polarization combination matrix is vectorized as [NodeB+45UEvert, NodeB-
45UEvert,NodeB+45UEhor, NodeB-45UEhor]. Here the notation NodeB+45UEvert refers to; from +45° slant element at NodeB
to vertically polarized element of UE (at nominal orientation).
For the 4 selected scenarios, SCM-A, SCM-B, SCM-C and SCM-D, the effective tap power that includes the effective
antenna gains at Node B and UE is tabulated below
Tap/mid- Delay [ns] Power, Ptap Node B spatial Polarization covariance matrix, Γ [4x4]
path [dB] correlation, α
3GPP
Release 7 118 3GPP TR 25.814 V7.1.0 (2006-09)
Ptap / 10
1 α ∗
Total per-tap covariance matrix: R = 10 ⋅ ⊗ Γ where the symbol ⊗ denotes the Kronecker product.
α 1
Tap/mid- Delay [ns] Power, Ptap Node B spatial Polarization covariance matrix, Γ [4x4]
path [dB] correlation, α
Ptap / 10
1 α ∗
Total per-tap covariance matrix: R = 10 ⋅ ⊗ Γ where the symbol ⊗ denotes the Kronecker product
α 1
3GPP
Release 7 119 3GPP TR 25.814 V7.1.0 (2006-09)
λ spacing, Laptop)
Table A.1.3-4 SCM-C (Urban Macro (high spread), 3-sector, 4λ
Tap/mid- Delay [ns] Power, Ptap Node B spatial Polarization covariance matrix, Γ [4x4]
path [dB] correlation α
UE spatial correlation β
1/1 0 -0.4616 + 0.5439i 0.5953 0.4047 0 0
1/2 12.5 0.00 0.0225 - 0.0595i 0.4047 0.5953 0 0
1/3 25.0 -2.22 0 0 0.5953 -0.4047
-3.98 0 0 -0.4047 0.5953
2/1 362.5 -1.86 0.2806 + 0.6476i 0.6134 0.3866 0 0
2/2 375.0 -4.08 0.0088 + 0.0602i 0.3866 0.6134 0 0
2/3 387.5 -5.84 0 0 0.6134 -0.3866
0 0 -0.3866 0.6134
3/1 250.0 -1.08 -0.1136 - 0.6818i 0.6090 0.3910 0 0
3/2 262.5 -3.30 0.0307 + 0.0555i 0.3910 0.6090 0 0
3/3 275.0 -5.06 0 0 0.6090 -0.3910
0 0 -0.3910 0.6090
4/1 1037.5 -9.08 0.6944 + 0.5043i 0.6430 0.3570 0 0
4/2 1050.0 -11.30 -0.0244 - 0.0028i 0.3570 0.6430 0 0
4/3 1062.5 -13.06 0 0 0.6430 -0.3570
0 0 -0.3570 0.6430
5/1 2725.0 -15.14 0.4072 + 0.5626i 0.6935 0.3065 0 0
5/2 2737.5 -17.36 0.0828 - 0.2378i 0.3065 0.6935 0 0
5/3 2750.0 -19.12 0 0 0.6935 -0.3065
0 0 -0.3065 0.6935
6/1 4600.0 -20.64 -0.7753 + 0.1776i 0.7535 0.2465 0 0
6/2 4612.5 -22.85 0.4194 - 0.2429i 0.2465 0.7535 0 0
6/3 4625.0 -24.62 0 0 0.7535 -0.2465
0 0 -0.2465 0.7535
1.
Ptap / 10
1 α ∗ 1 β ∗
Total per-tap covariance matrix: R = 10 ⋅ ⊗Γ⊗ where the symbol ⊗ denotes the
α 1 β 1
Kronecker product.
λ spacing, Laptop)
Table A.1.3-5 SCM-D (Urban Micro, 6-sector, 4λ
Tap/mid- Delay [ns] Power, Ptap Node B spatial Polarization covariance matrix, Γ [4x4]
path [dB] correlation α
UE spatial correlation β
1/1 0.0 -0.0907 + 0.1632i 0.5792 0.4208 0 0
1/2 12.5 0.00 0.0225 - 0.0595i 0.4208 0.5792 0 0
1/3 25.0 -2.22 0 0 0.5792 -0.4208
-3.98 0 0 -0.4208 0.5792
2/1 287.5 -3.57 0.0301 - 0.1586i 0.5792 0.4208 0 0
2/2 300.0 -5.79 0.0061 - 0.0051i 0.4208 0.5792 0 0
2/3 312.5 -7.55 0 0 0.5792 -0.4208
0 0 -0.4208 0.5792
3/1 200.0 -29.05 -0.5144 - 0.3812i 0.5792 0.4208 0 0
3/2 212.5 -31.27 0.0297 - 0.0078i 0.4208 0.5792 0 0
3/3 225.0 -33.03 0 0 0.5792 -0.4208
0 0 -0.4208 0.5792
4/1 662.5 -20.94 0.1275 + 0.0979i 0.5792 0.4208 0 0
4/2 675.0 -23.15 -0.0244 + 0.0029i 0.4208 0.5792 0 0
4/3 687.5 -24.91 0 0 0.5792 -0.4208
0 0 -0.4208 0.5792
5/1 812.5 -5.28 -0.0943 + 0.1609i 0.5792 0.4208 0 0
3GPP
Release 7 120 3GPP TR 25.814 V7.1.0 (2006-09)
Ptap / 10
1 α ∗ 1 β ∗
Total per-tap covariance matrix: R = 10 ⋅ ⊗Γ⊗ where the symbol ⊗ denotes the
α 1 β 1
Kronecker product.
The system simulation parameters for the micro cell scenario used for initial MIMO system level simulations are given
in Table A.2.1.1-4. The minimum set of micro cell simulation cases are given in Table A.2.1.1-2.
Other scenarios may, and higher velocities (e.g. 120km/h) shall be also verified.
3GPP
Release 7 121 3GPP TR 25.814 V7.1.0 (2006-09)
Total BS TX power (Ptotal) 43dBm – 1.25, 5MHz carrier, 46dBm - 10MHz carrier
Inter-cell Interference Modelling UL: Explicit modelling (all cells occupied by UEs),
3GPP
Release 7 122 3GPP TR 25.814 V7.1.0 (2006-09)
Table A.2.1.1-4 Micro-cell system simulation parameters for initial or early MIMO simulation results
Parameter Assumption
Outdoor to indoor Out-door to outdoor
Cellular Layout Hexagonal grid, 19 cell sites, 1 sectors per site
Inter-cell Interference Modelling UL: Explicit modelling (all cells occupied by Ues),
Minimum distance between UE and cell >= 10m (and minimum coupling loss of -53dB)
3GPP
Release 7 123 3GPP TR 25.814 V7.1.0 (2006-09)
Channel model for initial or early simulations Multi-Antenna Link level channel models (section
A.1.3)
It is intended to use SCM as defined in TR25.996 for MIMO simulations up to 5MHz and its extension (SCME) [R4-
060334] for MIMO simulations for higher bandwidths.
3GPP
Release 7 124 3GPP TR 25.814 V7.1.0 (2006-09)
VoIP DL and UL
Streaming DL and UL
Gaming UL
Transmitter 1 Antenna
3GPP
Release 7 125 3GPP TR 25.814 V7.1.0 (2006-09)
Transmitter 1 Antenna
Noise Figure 9 dB
BS antenna gain plus cable loss 14 dBi for micro, macro cell case
Noise Figure 5 dB
(P_OVHD)
HS-DSCH/HS-SCCH/DPCH
3GPP
Release 7 126 3GPP TR 25.814 V7.1.0 (2006-09)
Noise Figure 5 dB
BS antenna gain plus cable loss 14 dBi for micro,macro cell case
Pilot channel overhead Total time and/or power resources dependent on MA and
numerology are given or accounted for in simulation.
Control channel overhead Total time and/or power resources dependent on MA given
or accounted for in simulation
A.2.2.1 MIMO
In the evaluation of MIMO techniques for EUTRA MA candidates the following areas need to be aligned. It is
necessary to provide non-MIMO performance as a benchmark before or along with MIMO performance. Specific
MIMO schemes simulated for the work item phase should be accurately described.
3GPP
Release 7 127 3GPP TR 25.814 V7.1.0 (2006-09)
A.2.2.2 SDMA/Beamforming
More than 2 EUTRA Node-B antennas are likely needed to evaluate SDMA and Beamforming. Defining reference
EUTRA Node-B with 4 or more antennas is TBD.
1 Frequency reuse of a x b where ‘a’ is site reuse and ‘b’ is sector reuse.
3GPP
Release 7 128 3GPP TR 25.814 V7.1.0 (2006-09)
0.09
100
Traffic: Mod. ETSI - Web 2% outage
0.088 Browsing (user t-put <32Kbps)
Scheduler: PF
0.086
User t-put per Hz
Site-to-Site: 1.0km
0.082
At 2% outage point:
0.08 User t-put per Hz 5% CDF Point = 0.022
HSDPA: (5MHz FDD)
0.078 Receiver: MMSE
Codes: 15 - cch code ovhd
0.076 CCH: code, power, error ovhd
Common Chan: 20% power ovhd
0.074 150
0.072
0.35 0.4 0.45 0.5 0.55 0.6
Sector t-put per Hz
Figure A.2.4-1 – Example of User vs. Sector per Hz throughput with 5% CDF user throughput given at 2%
outage point.
0.03
IAT=2.0 seconds
CELL_DCH FACH: 2s
CELL_FACH DCH: 200ms
0.02 Traffic: Web Browsing
with TCP slow start Node-B: 15 codes available
Scheduler: PF
0.015
Channel: VA3/PB3
Site-to-Site: 1.0km
0.01
0
0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5
Sector Throughput per Hz
Figure A.2.4-2 – Example of 5% CDF User throughput vs. Sector throughput (per Hz)
3GPP
Release 7 129 3GPP TR 25.814 V7.1.0 (2006-09)
A.3.1 Downlink
Topic Aligned Value(s) - baseline for simulation
Resource block definition Localised resource blocks are baseline for full-buffer simulations.
Localised or distributed resource blocks can be used for VoIP
simulations.
The resource block (PRB and VRB) size is 25 subcarriers.
Data multiplexing (using LRB and DRB May be needed only for combined VoIP + full buffer or VoIP +
in same TTI) HTTP simulations.
Downlink reference signal structure Use TR25.814 agreement (include 2nd reference symbol as baseline
in link-level). Take appropriate overhead into account in system
simulations.
3GPP
Release 7 130 3GPP TR 25.814 V7.1.0 (2006-09)
Inter-cell interference randomisation Scrambling implicitly included (no impact on system simulations)
IDMA and frequency hopping not included in baseline
Inter-node B synchronisation Unicast (FDD): unsynchronised (not relying on methods that exploit
synchronisation)
Unicast (TDD) and Multicast: synchronised
Control signalling Overhead must be described and accounted for assuming at least
95% area coverage reliability.
A.3.2 Uplink
Topic Aligned Value(s) - baseline for simulation
Resource block definition Localized blocks are baseline for shared data channel.
Reference signal structure Account for overhead for system level simulations.
Assume FDM for data demodulation in link-level simulations
3GPP
Release 7 131 3GPP TR 25.814 V7.1.0 (2006-09)
Inter-cell interference randomisation Scrambling implicitly included (no impact on system simulations)
IDMA not considered in baseline
Frequency hopping on a TTI basis may be additionally considered in
the scheduler implementation
Control signalling Overhead must be described and accounted for assuming at least
95% area coverage reliability.
3GPP
Release 7 132 3GPP TR 25.814 V7.1.0 (2006-09)
Change history
Date TSG # TSG Doc. CR Rev Subject/Comment Old New
4-2005 RAN1 R1-050372 Initial TR skeleton presented for approval 0.1.0
#40bis
6-2005 RAN1 R1-050682 Basic transmission schemes and simulation assumption 0.1.1
#adhoc
8-2005 RAN1#42 R1-050698 Editorial correction 0.1.1 0.1.2
8-2005 RAN1#42 R1-050975 Raised to v0.2.0 following presentation of v0.1.2 at RAN1#42 0.1.2 0.2.0
10-2005 RAN1#42 R1-051029 Basic transmission schemes 0.2.0 0.2.1
bis
10-2005 RAN1#42 R1-051252 Raised to v0.3.0 following presentation of v0.2.1 at RAN1#42bis 0.2.1 0.3.0
bis
11-2005 RAN1#43 R1-051305 Duplexing, basic transmission schemes, multiplexing, enhanced 0.3.0 0.3.1
modulation schemes
11-2005 RAN1#43 R1-051569 Raised to v0.4.0 following presentation of v0.3.1 at RAN1#43 0.3.1 0.4.0
11-2005 RAN1#43 R1-051623 Reference signal structure for OFDM, Basic scheme for MC- 0.4.0 0.4.1
WCDMA, Initial simulation results, Multiplexing
11-2005 RAN1#43 R1-051264 MIMO 0.4.1 0.4.2
11-2005 RAN#30 RP-050729 Raised to v0.4.0 following presentation of v0.4.2 at RAN1#43 0.4.2 0.5.0
11-2005 RAN#30 Endorsed to v1.0.0 0.5.0 1.0.0
11-2005 Following by the decision in RAN#30 (RP-050909), scope is 1.0.0 1.0.1
updated
2-2006 Reference signal structure, multiplexing. Scheduling, HARQ, Link 1.0.1 1.0.2
adaptation is updated
2-2006 RAN1#44 R1-060363 Editorial correction 1.0.2 1.0.3
2-2006 RAN1#44 R1-060714 Raised to v1.1.0 following presentation of v1.0.0 at RAN1#42bis 1.0.3 1.1.0
2-2006 RAN1#44 R1-060763 Interference mitigation, Scheduling, HARQ, Link adaptation in UL, 1.1.0 1.1.1
MIMO etc.
3-2006 RAN#31 RP-060178 Raised to v1.2.0 following presentation of v1.1.1 at RAN1#44 1.1.1 1.2.0
3-2006 Editorial correction 1.2.0 1.2.1
3-2006 RAN1#45 R1-061110 L1/L2 control channel, MBMS. Cell search, RACH 1.2.1 1.2.3
bis
5-2006 RAN1#45 R1-061582 Raised to v1.3.0 following presentation of v1.2.3 at RAN1#45 1.2.3 1.3.0
5-2006 Performance evaluation, Cell search, RACH, MBMS 1.3.0 1.3.1
5-2006 RAN1#45 R1-061645 MIMO 1.3.1 1.3.2
5-2006 RAN1#45 R1-061647 Raised to v1.4.0 following presentation of v1.3.2 at RAN1#45 1.3.2 1.4.0
5-2006 R1-061650 Editorial correction 1.4.0 1.4.1
5-2006 R1-061651 Endorsed to v1.5.0 to be presented at TSG RAN #32 1.4.1 1.5.0
6-2006 RAN#32 RP-060347 Submitted for approval at TSG RAN #32 1.5.0 2.0.0
12/06/06 RAN_32 RP-060347 - - Approved as v7.0.0 to put under change control 2.0.0 7.0.0
29/09/06 RAN_33 RP-060496 0003 1 DL MIMO System Analysis 7.0.0 7.1.0
29/09/06 RAN_33 RP-060496 0004 1 DL and UL VoIP capacity characterization for E-UTRA 7.0.0 7.1.0
29/09/06 RAN_33 RP-060496 0008 1 Inclusion of Beamforming Results in TR 25.814 7.0.0 7.1.0
29/09/06 RAN_33 RP-060496 0010 1 Summary of downlink enhancement techniques over reference 7.0.0 7.1.0
LTE unicast performance
3GPP